Université Paris-Saclay Artificial Intelligence Informatics Master

[1st year (M1 UPSaclay)] [2nd year (M2 UPSaclay)] [EIT DSC master in Data Science]

AI informatics master program

Frequently Asked Questions

Kim Gerdes & Isabelle Guyon: Coordinators of the AI master program

with the help of François Landes (admissions and jurys) & Marc Evrard (internship coordination)

Alexandre Verrechia (UPsaclay) & Aurélie Lerasle (EIT digital): Administrative assistants

Enrollment 2020/2021 CLOSED

We received 825 applications in M1 (21 accepted) and 358 in M2 (28 accepted)

Congratulations to the new promotion!

[For fresh information: join the student Google group]

PLEASE WEAR MASKS WHEN YOU COME TO CLASS

The university mandates that masks be warn in all closed spaces and outside when distance >1m cannot be respected.

Contents:

Is this master for me?

Is this a one-year or a two-year program?

Two years: M1 and M2.

What is the entry level?

Bachelor in Computer Science, Statistics or Math (3 or 4 years) or French L3 completed (licence en informatique ou mathématique+informatique).

Are courses taught in English or French?

The language of instruction of all mandatory classes is English. However:

- the instructors are (mostly) French, so you should be ready to get sometimes relatively “broken English”. Also, some of the slides or teaching material might not be translated (yet).

- some options from other parcourses are taught in French.

What skill will I learn?

- Applied statistics and optimization.

- Handling big data.

- Supervised, unsupervised, and reinforcement learning, theory and practice.

- Predictive modeling and data visualization with Python, scikit-learn, Pytorch, Tensorflow.

- Cutting edge AI modeling techniques, including Deep Learning.

- Applications of machine learning to computer vision, voice recognition, text processing, natural language processing.

- Writing project proposals and communicating results in writing and orally.

See the provisional course list. Each module is a "teaching unit", in French Unité d'Enseignement or UE], worth 2.5 ECTS.

How many ECTS do I earn from this program?

120 ECTS (60 each year). In total (see preliminary directives), you will have to earn:

- 70 ECTS by taking classes from the AI program,

- 20 ECTS by taking classes common to all, or borrowed from other master programs at UPSaclay ((languages, communication, entrepreneurship, research and development, conferences or seminars, etc.)

- 30 ECTS from the last semester 5-6 month internship.

What careers does this open to me?

Students who have followed this program will have solid theoretical and practical training. Projects and internships prepare them to enter the workforce. This master's degree opens towards research and teaching, as well as to careers in the industrial sector or services. Some of our graduates

end up working in startups or integrate R&D departments of companies such as Thales, Orange, HP, IBM Research, Yahoo, Google, Facebook.

Can I apply if I am NOT a Computer Science (CS) major?

Yes, but you should meet the prerequisites. CS, Math, Statistics, or Physics majors are probably best prepared.

What are the prerequisites?

Good bases in mathematics and computer science:

- Probability and statistics

- Linear algebra

- Differential and integral calculus

- Scientific programming

- Data visualization

Will there be entry-level M1 classes, to help me fill some of my "missing prerequisites"?

There are refresher classes (the "PRE" modules, and TC0), but not sufficient for students who have not taken such classes at all as undergraduates. Your grades in MATH in particular will be an important criterion of selection.

- PRE1: Applied statistics

- PRE2: Mathematics for data science

- PRE3: Datacomp

- PRE4: Scientific programming

- TC0: Machine learning 1

Where is UPSaclay, is it in Paris?

No, it is in the outskirts of Paris, about 1 hour away by public transportation.

For people who like outdoors, hiking, sports, it is really nice. You can get to Paris easily and enjoy theater, restaurants, monuments, and museums. Locally, there are arts-and-craft, theater, sports, and other clubs, and limited local shows and exhibitions.

Do students have the possibility of starting a Ph.D. thesis straight after graduating from M2?

In France, to start a PhD, you need to find an advisor first. To get a scholarship, you need to write a proposal and get letters of support. However, there are some advisors that already have funding and advertise for an open PhD position.

Should I apply for UPSaclay or EIT DSC?

It depends. The UPSaclay degree is a 2-year degree at UPSaclay (in France). The EIT DSC degree is a European degree in which you do 1-year in one university and the other in another one. You would do one of the two years at UPSaclay, for instance, and the other elsewhere in Europe. The classes are mostly the same, but the EIT program has extra "entrepreneurship" training.

Can I apply to UPSaclay even though I am an international student and the AI master in not marked [here] as an "international master"?

Yes!

What are the tuition fees?

The differ for UPSaclay and EIT Digital. The UPSaclay degree is relatively inexpensive, compared to EIT fees . There is an extra FEE for non EU citizens.

COVID-19 crisis: What will happen to the selected students if the 2020-2021 academic year begins before the borders open?

All courses will be available to follow remotely.


Can I apply and how?

Applications for 2020/2021 are over

Where and when can I apply?

- If you are a student from one of the countries listed on [this page], you must first go through the Campus France process.

- then go [HERE] for UPSaclay (application period from 01/05/2020 to 01/07/2020) -- download instructions. Contact support.

- or [HERE] for EIT DSC (applications for 2020 have ended).

Can I apply to multiple masters or parcourses?

Yes

What will I need to provide with my application?

- Curriculum Vitae (in any format)

- Letter of motivation:

* Describe a personal experience that convinced you to pursue in AI studies.

* Which classes you enjoyed in the past could be relevant to AI

* What are your favorite AI topics ?

* How do you see your future career as an AI graduate

- Certified transcripts of grades for all diplomas since high-school.

- "Fiche de choix" (selection sheet), see below.

- Dossier VAPP (optional, see below).

- Recommandation letter or internship evaluation (recommended).

Note: for international students, you are asked to provide 2 contact persons to serve as reference to recommend you. This is redundant with "Recommandation letter or internship evaluation". You can either upload the recommendation letter manually (yourself) in this section, or skip it. Your references will be contacted anyways separately to provide a recommendation.

Should the application be in French or in English?

Either way.

I submitted my candidature but forgot something, how can I add it?

You cannot, sorry. You must destroy your first candidature and create a new one!

I destroyed my application accidentally can you revert this or accept my files by email?

No. You need to create a new account and re-enter everything on the platform by hand.

Due to special circumstances (e.g. COVID-19 epidemic) I had to telecommute for my internship, should I mention it?

You can add internships performed while telecommuting in your work experience, as well as other work experiences performed during the confinement, as long as you could perform your duties.

Where is the "Selection Sheet" ("Fiche de choix"), and what is this good for?

It is "somewhere: on the website, but if you cannot locate it:

- Entering M1

- Entering M2

Use it to rank order CS mater specializations by preference. Or use this:

- NEW MODEL

Is the "M1 machine learning" the same thing as "M1 artificial intelligence"?

Yes, there seems to be an incoherence on the platform. If you see that, it is OK.

Can I enter directly in the second year (M2 level)?

We do not recommend it, unless you have completed a 4-year Computer Science Bachelor degree or equivalent, with a strong emphasis on Machine Learning/AI/Data science. The classes you should have taken should include most of the M1 classes of our program, and at least all PRE classes, TC0-3, and OPT4. In exceptional cases, you may also apply to enter at the M2 level if you interrupted your studies more than 2 years can justify of continuous education classes and/or professional experience that is equivalent to the M1 classes we offer. In this case, you should prepare a special application (VAPP=Validation des Acquis Professionnels et Personnels).

What are the conditions to attend the M2 level if I am a third-year student of an engineering school (double cursus)?

You may apply, but take into account the prerequisites. You may have to follow M1 level modules, e.g. 2 PRE classes to catch up and be allowed to follow more advanced classed. At the M2 level, two PRE classes count for only one module (2.5 ECTS). Check with your school the number of ECTS you need to graduate. As of last year, the rules were:

- The ENSIIE students must follow at least 10 modules instead of 12 in S3 (the 6 TC and 4 OPTs in theory). The students earn 5 ECTS of the ENSIIE. If they take more modules in the M2 program, only the 10 best results count in the average. In S4 they are exempt from FVE and English, unless they wish.

- The ENSTA students must follow at least 7 modules instead of 12 in S3 (3 TC and 4 OPTs in theory). The students earn 12.5 ECTS from ENSTA. If they follow more modules in the M2 program ,only the 7 best results count in the average. In S4 they are exempt from FVE and English, unless they wish.

- The CentraleSupélec students must follow at least 4 modules instead of 12 in S3 (2 TC and 2 OPTs in theory). The students earn 20 ECTS from CentraleSupélec. If they follow more modules in the M2 program, only the 4 best results count in the average. In S4 they are also exempt from FVE and English, unless they wish.

- The Télécom students must follow the M2 curriculum in S3, similarly to the "classic" M2 IA cursus. In S4 they are exempt from FVE and English, unless they wish.

Do I need to include a "Dossier VAPP"?

Not necessarily. Only if you cannot justify of the right academic profile and you want other personal or professional experiences to be taken into account. See VAPP instructions. This Dossier VAPP can be useful at the M1 or M1 entry level, if you interrupted your studies more than 2 years, and want to capitalize on personal or professional experience to compensate for missing academic credentials.

As a foreign student, I cannot see my diplomas in the list, what should I select?

- Baccalauréat ou équivalent:

* Equivalence BAC diplôme ou titre étrang.

* Equivalence BAC

- Cursus précédent:

* Bachelor of Science

Do I need to submit my high-school grade transcript?

No.

I am a foreign student, do I need to complete the Campus France Procedure (CEP) and when?

Yes, if you come from one of the following countries:

Algérie, Argentine, Bénin, Brésil, Burkina Faso, Burundi, Cameroun, Chili, Chine, Colombie, Comores, Congo, Corée du Sud, Côte d'Ivoire, Djibouti, Egypte, Etats-Unis, Gabon, Guinée, Haïti, Inde, Indonésie, Iran, Japon, Koweit, Liban, Madagascar, Mali, Maroc, Maurice, Mauritanie, Mexique, Nigeria, Pérou, République démocratique du Congo, Russie, Sénégal, Singapour, Taïwan, Tchad, Togo, Tunisie, Turquie, Vietnam.

See the full updated country list.

As soon as you are admitted, you must go through the procedure, if you are not already studying in France and need a VISA. These FAQ should help you.

Can I get a scholarship to help me pay for my studies?

- If you applied to the EIT program, you may be selected for a merit-based scholarship by the EIT program, depending on your rank in the application review process. EIT grants cannot be cumulated with Paris-Saclay grants or Campus-France grants.

- There are several merit-based scholarships from Paris-Saclay and Campus-France (see below). We must submit the request on your behalf.

- For Paris-Saclay grants, you must submit your application to the master program first!

- Do not wait for the application deadline of June 30 BECAUSE WE MUST REVIEW AND SELECT CANDIDATES before MAY 21.

- Please submit your application no later than MAY 15, 2020 and contact us by email to request a scholarship.

- If you miss the deadline of May 15, but it is BEFORE May 31st, contact us, in case we don't have many candidates for the Digicosme scholarship.

1) University Paris-Saclay international scholarships:

- If you are an international student, you may be contacted to receive a merit-based grant of IDEX Program "Investissement d’Avenir 3" [details][cadrage April 11].

- There are one or two-year scholarships of 10 000 € per student per year.

- Candidates selected for this grant will receive an email and instructions how to apply (they cannot apply directly).

- A prerequisite is to have been accepter to the master and pre-selected, hence please submit your application to the master program first and contact us by MAY 15, 2020.

- Your deadline for submission (once you are pre-selected) is MAY 25, 2020 (recommendation letters submitted should be received before MAY 27).

- Evaluation criteria:

* Academic level

* Personal project

* Motivation to pursue a PhD thesis (for M2 students).

2) Labex Digicosme scholarships:

- There are only three Digicosme grants for our program of 12 000 € per student per year.

- Submit a request letter to Isabelle Guyon, by MAY 15, 2020, including

* a copy of your ID (passport or identity card).

* Family name:

* First name:

* Email:

* Phone (please specify in which time zone you are and what is a convenient time to contact you):

* Gender:

* Citizenship:

* Current university/school:

* Highest degree and major (completed by start of this master degree):

* Date of highest degree (or anticipated date of completion):

* Spoken/written French level: None/Beginner/Good/Fluent

* Spoken/written English level: None/Beginner/Good/Fluent

* The name and contact information (email and phone) of two references.

- For the rest of the information, we will use your application. So, please submit your application to the master program first!

- This is a merit-based grant, but in case there are many applicants, please also justify of your needs is your request letter.

Exceptional candidates will be proposed for receiving a grant to the Labex Digicosme jury by Isabelle Guyon and Aurélien Decelle. They will add to the application of the student a reasoned opinion justifying the excellence of the student's training in relation to the selected AI master program and themes of the Labex DigiCosme. Criteria will include the quality of the institution of origin, recommendations, and direct discussions with the candidate. An important element will be the potential to pursue a PhD thesis.

If you get a grant in M1, the grant is renewable 1 year if you are accepted in M2 the following year.

3) Eiffel international scholarships:

- The Eiffel Program is reserved for non-French nationals. Candidates with two nationalities, one of which is French, are not eligible.

- Candidates must be 30 or under.

- Candidates must submit their application to the persons responsible for the master, NOT to the Eiffel program directly.

- The Eiffel scholarship consists of a monthly allowance of €1,181 (a maintenance allowance of €1,031 plus a stipend of €150) and other benefits. It can be obtained for one or two years.

- For this year the deadline is past (was in january 2020). Keep this in mind for 2021/2022 [see instructions from Université Paris-Saclay]. Contact: Mme. Julie HERISSON.

4) Charpak international scholarships:

- The Chapak program is reserved to Indian nationals.

- It includes a living allowance of 700 euros per month, a tuition waiver, and other benefits.

- It is conditioned on attaching a letter of acceptance to our master, which we cannot deliver until July or so, for first time entry students. So this will concern mainly M2 students, presumably.

- For this year the deadline is past (was in april 2020). Keep this in mind for 2021/2022.

5) Other scholarships

- Consult the directory of scholarships on campus france.

As an M1 student, can I get a grant to study abroad?

Yes, this is called a "bourse de mobilité sortante". For this year the deadline is past (was in march 2020). Keep this in mind for 2021/2022.

When will I know whether I am accepted?

Within 2 months of depositing your files (validated and submitted by you).

What are the admission criteria?

The jury will consider all applications and rank them according to: reputation of undergraduate degree, grades, pre-requisite classes taken, past job experiences, recommendations, and motivation. Everything counts.

How many students are admitted?

- In M1: 15 to 20 students, EIT included.

- In M2: 20 to 25 students, EIT included.


How do I register?

==> You first step: General information on registration

==> For questions not covered below, see the University Registration FAQ

==> For help from other fellow students: register to the Google group of the master.

Where can I get help?

Alexandre Verrecchia

Secrétaire pédagogique master IA

Bât. 640-PUIO, Bureau B206

Université Paris-Saclay

91400, Orsay, France

WARNING: The administrative help will be on vacation 25 juillet au 23 août

When is the deadline to confirm my application?

We are not aware of a specific deadline, but you must start the registration process as soon as possible because of the lengthy administrative process and the difficulty of finding lodging. By courtesy to other students on the waiting list, please confirm as soon as possible.

How to I fill out registrations forms?

- Go to the registration site

- Select Université Paris-Saclay, UFR des Sciences

- New students should go through IA-PRIMO.

- What is "Année du bac"? This is the last year of high school when you passed your final high school exam (baccalauréat in France)

What is the INE number and how to I get it?

This is a student registration number. You get it when you pay the CVEC (Contribution Vie Etudiante Campus)

What is the deadline to register?

On-line registration is open July 8-23 then August 24 to September 9.

Can I get a certificate of acceptance?

Yes email me.

How do I get medical insurance?

Register to get a social security number. It is also recommended to get a private medical insurance for the first 4 months because the French administration is notoriously slow. See this information page.

How to I get liability insurance?

Liability insurance (assurance de responsabilité civile) is mandatory in France. You usually need to get one when you rent. See also this information page.

Check on the internet "assurance responsabilité civile étudiant" to get various options.

What tuition fee do I have to pay?

The tuition fees are 243 €. The students also have to pay Contribution Vie Etudiante Campus (CVEC - 92 €).

EIT students do not pay CVEC in our university.

How can I get accommodations?

See list. Check distances to your study location (PUIO) and transportation means.

Email contact.

Does the univesity help me get a guarantor for my lease?

Check VISALE in the help page.

What is the university address?

Pôle Universitaire D'Ingénierie D'Orsay (PUIO)

Université Paris-Saclay (Bât. 640)

91400, Orsay, France

[see map]

Who can help me there with administrative issues?

Alexandre Verrechia

PUIO Bâtiment 640, Bureau B206

Open:

8H45-16h15 Tuesday and Thursday

8h45-17h15 Wednesday

8h45-15h15 Friday

When is the start date of school?

September 7, but there might be an orientation meeting the week before. You should be ready for school beginning of September.

Will I attend courses online or on the campus during the first semester?

On-line courses will be available for students who cannot physically attend.

University guidelines are [here]. Do not worry that the [AI] master is not mentioned in the list of masters providing on-line classes. We WILL provide on-line classes.

Wearing masks when entering the campus buildings and during classes in mandatory.

What is the yearly calendar (with vacations, etc.)?

See this.

Due to the Covid-19 pandemic, all international mailing services are not properly working. Do I still need to mail hard-copies of all documents to complete my registration?

YES.

- mail hard copies (translated in French by a certified person; you can find on-line services that do that)

- keep proof that you mailed the documents

- keep all originals and bring them with you.

"Les primo-entrants via IAPRIMO devront encore renvoyer leur récapitulatif et pièces justificatives par courrier à :

Université Paris-Saclay - UFR Sciences - service Scolarité bât. 336 – campus d’Orsay - 91405 Orsay cedex"

Who to contact to get a physical appointment if you are a foreigner that must pay dues "physically"?

Get an appointment by writing to :

scolarite-licencesmasters.sciences@universite-paris-saclay.fr

or call +33 (0)1 69 15 70 68.

Where to get more information?

HERE


How are classes organized?

During the COVID crisis and until further notice, instruction will be provided via teleconferences (or with pre-recorded material) and homework that can be submitted without physically attending classes. On-site tutorial sessions with the instructors may be provided for support (depending on the instructors). Wearing masks when entering the campus buildings and during classes in mandatory.

From 2020-2021, there will now be a global schedule system between Masters at Université Paris-Saclay, with semesters divided into 2 equivalent periods:

T1 / T5 : Sept. 7 to Oct. 23

T2 / T6 : Nov. 2 to Dec. 18

T3 / T7 : Jan. 4 to Feb. 19

T4: Mar. 1 to Apr. 16

See yearly schedule or EDT (Emploi Du Temps). Students can select UEs in other Masters of the mention "informatics". For the IA Master, it is planned that the courses will take place on Thursday and Friday with approximately 4 courses per period. Each course takes 1/2 day (3.5 hours) and courses last 7 weeks (including 1 exam week, if there is a final exam). Each UE will be worth 2.5 ECTS (credits). Wednesdays will be reserved for "soft skills" such as communication, conferences, language courses ...

Travail d'Etude et de Recherche (TER) = Small research internship

This teaching unit consists in carrying out a short internship of initiation to research in a team of a research laboratory. The rate is 1 day (or two half-days) per week. If the internship is longer than 2 months it must be remunerated.

Teacher responsible for I&E basics/BDlabs (M1 EIT) is:

Guillaume DION

Teacher responsible for I&E basics (M2 EIT ) is:

Alvaro PINA STRANGER

The students are expected to validate 120 ECTS over 2 years:

- In M1 (60 ECTS):

  • Bloc 1: 20 ECTS through 8 mandatory UEs for the AI Master (including a project).

    • Bloc 2: 20 ECTS through 8 optional UEs, which may include AI Master UEs or other informatics UEs.

    • Bloc 3: 20 ECTS through "soft skills" UEs (such as languages; also called UE d'ouverture) and a TER internship

- In M2:

    • Bloc 1: 15 ECTS through 6 mandatory AI Master UEs

  • Bloc 2: 7.5 ECTS through 3 optional UEs, which may include AI Master UEs or other informatics UEs.

    • Bloc 2: 7.5 ECTS through "soft skills"

    • Bloc 4: 30 ECTS internship, in the second semester

WARNING: there are still a few inconsistencies because the program is new and it is very difficult to get the administrators to update [this site]. Please bear with us.

What courses can I choose from?

We prepared for you surveys to be filled out to choose your classes. Please turn in the answers as soon as possible and NO LATER THAN AUG 24.

M1: https://forms.gle/AS4hXXo7mZaW5DxN7

M2: https://forms.gle/thWSLXJKcKTeRPGGA

You must take some mandatory classes and choose from optional classes. Some classes from other parcourses may also be available, see classes underlined in the provisional weekly schedule or weekly EDT.

If you do not have the right prerequisites, contact the instructors.

M1 classes, 60 ECTS

Classes in bold are mandatory. Others are optional. Among the bold classes, you may take either green (ML path) or brown (NLP path) classes in a given period.

AM = morning

PM = afternoon

ECTS = credits

(*) Choose the green (ML) or brown (NLP) class.

(*) Language classes:

--> English lessons start on Monday September 9 in the house of languages - Bat. Eiffel - 3rd floor in front

the small amphitheater at 8.45 am.

--> FLE French classes for non-French speaking students (Français Langue Etrangère):

The students will be placed in groups of level A1 to B1/C1.

Placement test on Thursday, September 24, 16h30, then 12 weekly lessons of 2 hours on Thursday evening (Eiffel Building 3rd floor, specific rooms will be posted in room LC 353).


M1 Artificial Intelligence class list

refresher

Refresher classes

PREparatory classes, 2.5 ECTS each. Mandatory for all M1 students. Prerequisite for M2 students who have not taken similar classes. If you cannot take them, study on your own, see: How do I get ready?

TC

Foundational classes

Tronc commun (TC) “classic classes", 2,5 ECTS each.

Among the bold classes, you may take either green (ML path) or brown (NLP path) classes in a given period.

growth

Growth classes

Formerly all OPTional classes, though some in bold are now mandatory :-) 2.5 ECTS each.

    • OPT4: DEEP LEARNING (Caio Corro, Michèle Sebag) -- With ML2 as prerequisite

    • OPT8: HISTORY OF AI (Kim Gerdes)

    • OPT9: DATA CAMP (details) (Marc Evrard, Adrien Pavao) Challenges in machine learning; participate to a challenge -- With PRE1 and PRE2 as prerequisite

    • OPT13: INFORMATION THEORY (website from a previous similar class). Théorie de l'information (Guillaume Charpiat) -- With PRE1 as prerequisite

    • DS: DISTRIBUTED SYSTEMS. UE from Data Science parcourse.

    • HCI: INTERACTIVE MACHINE LEARNING. UE from Human Computer Interface parcourse.

project

Projects and practical experiences

    • Summer school (école thématique)

    • Internship (stage)

    • Project A (Website from previous years): Create a mini-challenge that other students will solve. Team work in teams of 5-6 people. (Isabelle Guyon et Kim Gerdes) -- With ML1 and PRE4 (or equivalent) as prerequisite

    • Project B (Website from previous years): Resolve a mini-challenge. Team work in teams of 5-6 people. (Kim Gerdes, Marc Evrard)

    • TER: Travail d'Etude et de Recherche. Personal work supervised by a member of the master's teaching team. This work can take the form of a state of the art on a given scientific subject and / or the implementation of state algorithms for application on a given problem. This work is not remunerated.

Other "soft" skills

    • Languages: French or English for non native speakers (Roselyne Debrick).

    • I&E: Innovation and Entrepreneurship.

    • FVE: Research and Development training, mutualised with avec MIAGE d'Orsay. Formation à la Vie en Entreprise (Nadège Taillard, Alexandre Kaminski, Frédérique Blondel)

M2 classes, 60 ECTS

(*) Choose the green (ML) or brown (NLP) class.

BLOC 4: INTERNSHIP, 30 ECTS

M2 Artificial Intelligence class list

TC

Foundational classes

Tronc commun (TC) “classic classes", 2,5 ECTS each. Both are mandatory.

growth

Growth classes

Formerly all OPTional classes, though some are now mandatory :-) 2.5 ECTS each.

Among (mandatory) bold classes, you may take either green (ML path) or brown (NLP path) classes in a given period. Green and red classes can also be taken as options.

OPT1: GRAPHICAL MODELS (details) (website from last year) (Francois Yvon) Modèles graphiques pour l’accès à l'information à grande échelle -- With TC4 as prerequisite

OPT2: IMAGE UNDERSTANDING (details) (website from last year) (Isabelle Bloch) Interprétation d'images -- With TC1 as prerequisite

OPT3: REINFORCEMENT LEARNING (website from last year) (Michele Sebag) Apprentissage par renforcement -- With TC1 as prerequisite

OPT5: VOICE RECOGNITION AND AUTOMATIC LANGUAGE PROCESSING (selection of papers from last year) (Kim Gerdes, Marc Evrard)

OPT6: LEARNING THEORY AND ADVANCED MACHINE LEARNING (website from last year) (Antoine Cornuejols) Apprentissage avancé et théorie -- With TC1 as prerequisite

OPT7: ADVANCED OPTIMIZATION AND AUTOMATED MACHINE LEARNING (Isabelle Guyon, Zhen Xu) (Anciennement Optimisation avancée, nouveau cours) -- With TC2 as prerequisite

OPT 10: IMAGE MINING (details) (website from last year) (Antoine Manzanera) ancien Indexation d'image et recherche par le contenu -- With TC1 and 6 as prerequisite

OPT 11: DEEP LEARNING FOR NLP (website from last year) (Caio Corro) Natural Language processing -- With OPT4 as prerequisite

OPT 12: TEXT MINING AND CHATBOTS (Anne Vilnat) -- With TC3 and 6 as prerequisite

OPT 14: MULTILINGUAL NATURAL LANGUAGE PROCESSING (details) (François Yvon) -- With TC4 as prerequisite

Soft skills

  • Communitation

  • I&E: Innovation and Entrepreneurship.

  • Conferences

project

Internships

5 to 6 month internship in a research lab or a company. (Marc Evrard, coordination)

Your primary source of information should be the site of Marc Evrard.


How do I get ready for the AI master program?

If you want to prepare yourself over the summer, there are several classes that we recommend, which can help you:

- At M1 entry level, prepare for striving in the program

- At M2 entry level, replace missing prerequisites:

We HIGHLY RECOMMEND at least reading the THREE CRASH COURSES highlighted in red.

Advanced students will be interested in the Machine Learning Summer School

Replace, review, or prepare for:

  • PRE1: Applied statistics

[CRASH COURSE] Crash Course on Basic Statistics, Marina Wahl (28 pages + questions)

[ON-LINE BOOK w. exercises] Computational and Inferential Thinking, By Ani Adhikari and John DeNero

[BOOK] Think stats, AB Downey.

[BOOK] Statistics in a Nutshell. Sarah Boslaugh and Paul Andrew Watters

[BOOK] All of statistics: a concise course in statistical inference. Larry Wasserman

[CHEATSHEET] Harvard cheatsheet in statistics

  • PRE2: Mathematics for Data Science

A) LINEAR ALGEBRA

[CRASH COURSE] Linear Algebra Review and Reference. Zico Kolter and Chuong Do (26 pages)

[SHORT COURSES] Linear Algebra, Khan Academy

[COURSE] Mathematics for Machine Learning: Linear Algebra, David Dye, Coursera (19 hours)

B) CALCULUS

[CRASH COURSE] The matrix calculus you need for deep learning, T Parr, J Howard (33 pages)

[COURSE] Mathematics for Machine Learning: Multivariate Calculus,

Samuel J. Cooper, Coursera (19 hours)

[BOOK] Matrix Computations. Gene Golub and Charles van Loan

    • PRE3: Datacomp1

[COURSE] Databases and SQL for Data Science. Rav Ahuja

      • PRE4: Scientific programming

[COURSE] Data Analysis with Python. Joseph Santarcangelo

  • TC0 and TC1: Machine Learning

[TUTORIAL] An introduction to machine learning with scikit-learn

[TOTAL BEGINNER COURSE] Introduction to machine learning, Sebastian Thrun, Katie Malone, Udacity (10 lessons)

[BEGINNER COURSE] Machine Learning, Andrew Ng, Coursera (54 hours)

[BOOK] The Hundred-Page Machine Learning Book. Andriy Burkov.

[BOOK] Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow. Aurélien Géron.

[ADVANCED COURSE] How to win a data science competition. Dmitry Ulianov, Coursera, (4 weeks); Excellent summer program!

  • TC2: Optimization

[COURSE] Introduction to optimization, by the instructors of TC2: Anne Auger and Dimo Brockhoff

  • OPT7: Advanced optimization

[COURSE] Advanced optimization, by the instructors of OPT7: Anne Auger and Dimo Brockhoff

  • OPT13: Information Theory

[COURSE] Information Theory, Inference, and Algorithms, David MacKay

How will classes be organized during Covid?

The university has issued guidelines, which are [summarized here]. See also this "cheatsheet".

Given government recommendations and the university policy, last update aug 28, as well as the fact that many students of the AI parcourse are from abroad and have difficulties obtaining a visa, at least in the first semester, theoretical classes and some exercise classes will be taught remotely. It will be possible for students who cannot physically come to campus to study entirely remotely. The instructors will communicate the modalities to the students.

ALL students are responsible to get equipped with a personal compter an Internet connection allowing them to attend videoconferences and perform homework remotely. If you have difficulties getting proper equipment, please contact us.

Although your presence on campus is not strictly required, meeting with your instructors, at least occasionally, for question sessions and/or practical work, and working with your fellow students may enhance your learning experience. We encourage you to do so, but to respect strictly university regulations:

- PLEASE WEAR MASKS WHEN YOU COME TO CLASS.

- The university mandates that masks be warn in all closed spaces and outside when distance >1m cannot be respected.

If you cannot (or do not want to) respect these guidelines, DO NOT GO TO CAMPUS. To get a free cloth mask, please contact us.

- International students must get a medical examination, contact medecine preventive. Follow advice to international students.

Students are encouraged to get in touch with each other and self-organize. To that end, we have set up a Google group. Safely meet to study together in small group observing social distancing and wearing masks. On place to meet is La maison des étudiants de Paris-Saclay. Preferably meet outside when the weather permits.

Do not wait until you get anxious, depressed, or demotivated to ask for help, please contact us. We conducted a survey last semester on studying during confinement. Here are some tips other students want to share with you to study effectively:

    1. Attend classes regularly

    2. Get fast and reliable Internet

    3. Do homework regularly

    4. Use on-line resources

    5. Communicate on-line with other students

    6. Do homework together with other students

    7. Alternate studying and doing physical exercise

    8. Take regularly little breaks

Besides, take good care of yourself:

  • Eat well

  • Sleep well

  • Exercise

  • Drink water regularly

  • Go out in the sun

And do not neglect your social life:

  • Have regular video-conferences

  • Check on your friends and family

  • Cook or eat together (even remotely)

  • Play games together (even remotely)

Nor your hobbies:

  • Learn something new

  • Talk to someone

  • Do some sports

  • What TV, videos, or read

  • Cook a nice meal

  • Do some arts and crafts

Get other ideas from the survey.


Where are my classes?

Latest information October 31, 2020: Due to the Covid situation, all classes will be held on-line.

Schedule == Emploi Du Temps == EDT

M1 schedule

M2 schedule

Overall on-line schedule

We are now mostly using Ecampus. If you still do not have access, please contact us.

M1 classes on Ecampus

M2 classes on Ecampus

M1 second period (T1) classes (nov-dec 2020):

TC0: Machine Learning 1 [INFO][external site][ecampus]. Fri PM

TC2: Optimization [external site][ecampus]. Thu PM

OPT8: History of AI [external site][ecampus]. Fri AM

OPT9: Data Camp [external site][ecampus]. Thu AM


M2 second period (T6) classes (nov-dec 2020):

OPT1: Graphical models [external site] [ecampus]. Fri AM

OPT3: Reinforcement learning [OLD external site][ecampus]. Thu AM

OPT5: Voice recognition [external site][ecampus]. Fri PM

OPT11: Deep learning for NLP [external site][ecampus]. Wed AM


Older information

Slides of presentation of the master

Location of most classes (unless on-line) BRING YOUR MASKS

Pôle Universitaire D'Ingénierie D'Orsay (PUIO)

Université Paris-Saclay (Bât. 640)

91400, Orsay, France

[see map]

For on-line classes

- Computer rooms D206 and D207 are open every day from 8 a.m. to 7 p.m.

- Room E 213 is reserved for students who want to do homework together.

Students will either have Internat access with the provided computers, or can use their cell phone or their personal computers via the WiFi.

No microphone is provided to them, please come equipped your own headset / microphone, because several courses will use the same rooms.


Language classes

English lessons start on Monday September 9 in the house of languages - Bat. Eiffel - 3rd floor in front

the small amphitheater at 8.45 am.

French lessons: Placement test on Thursday, September 24, 16h30, then 12 weekly lessons of 2 hours on Thursday evening (Eiffel Building 3rd floor, specific rooms will be posted in room LC 353).

Address: CentraleSupelec, Bâtiment Eiffel 3 rue Joliot-Curie F-91192 Gif-sur-Yvette Cedex

M1 first period (T1) classes (sep-oct 2020):

[AI] PRE1: online-class, website. For those who want to attend physically, go to PUIO E107

[AI] PRE2: online-class, website

[AI] PRE3: online-class, website

[AI] PRE4: online-class, website

Options from other parcourses:

[DS] Bases de données avancées 1: PUIO D201-202, Wednesday afternoon, on-line class , at least the first week

[PDCS] Modélisation et optimisation des systèmes discrets: PUIO E212

[PDCS] Algorithmique parallèle (in French) website

[ANO] Optimisation dans les graphes (in French)

M2 first period (T5) classes (sep-oct 2020):

[AI] TC4: online-class, website

[AI] TC5: online-class, website

[AI] OPT2 and 10: online-class, website

Options from other parcourses:

EIT Innovation and Entrepreneurship 1

EIT Career seminar [WARNING: EIT must take this class; schedule conflict with TC5: replace by one of the image classes OPT2 or 10]

[ANO] IoT: PUIO E 105

[HCI] Design Project 1: PUIO E 105

[PDCS] Initiation au calcul quantique (in French) website

[PDCS] Calcul haute performance (in French) website

Note: students who Career Seminar 2 in T5 MUST also take Career Seminar 2 Project in T6.


Internships, employment, and PhD theses

Your primary source of information should be the site of Marc Evrard.

What should I do to find an internship?

- Check thee list of proposed internships Internship offer [AI] informatics master (Responses) which will keep growing.

- We encourage you to also search on your own. Companies such as Microsoft, Google, Facebook, Thales, Safran, RTE, Datadog, Dassault, Schlumberger, BNP Paribas, Société Générale, Total, IFP, Renault, Guerlain, 4Paradigm, Airbus, Quantcube, Essilor, Unilever, NukkAI, Onera, GE healthcare, Thomson, and research labs at Paris-Saclay (INRIA, CNRS, CEA) have been offering internships in the past to our students.

- Subscribe to various mailing lists, such as Bulle-i3 (general CS), ATALA (NLP), or Parole (Speech), and TAU (Machine Learning) where you'll find internship adds.

- Get in touch with teachers or researchers whose fields you're interested in.

Please start applying now, and get back to Marc Evrad by fill up THIS FORM.

November 15, 2020

with a list of 2 to 4 internships you have selected (in order of preference), in the proposed list Internship offer [AI] informatics master (Responses). But do not wait until then to start applying!!! If you find an internship NOT listed yet, please fill the application form with the required information, and add your name in the field :

"Name of a student from our program that you want to direct this internship to”.

You can also follow the recommendations of the university.

How do I find a PhD position and a scholarship?

Check ADUM

How long should an internship be?

- M1 students: One month full time (e.g. during vacations) or two months part time (one day per week at least).

- M2 students: Five to six months.

Are internships graded?

- M1 students: No, they are PASS / NO PASS, you just need to fulfill the requirement.

- M2 students: Yes.

What is the difference between and internship and a TER (M1 students)?

- Internships:

* can be done in a company or a research lab

* can be done full time during vacations (1 month) or at least 1 day/week for 2 months during school time

* are NOR graded

- TER (Travail d'Etude et de Recherche):

* should be done in a research lab at the university

* at least 2 1/2 days per week for 2 months

* will be graded

Both can be remunerated. A minimum is guaranteed by law.