International Bachelor Ygrec Data Science by Design - ENLIGHTENMENT SESSIONS

Anne-Laure GALLINE CASTEL

She has a Master 2 in business Law (Université Paris II - Panthéon-Assas), a Master 2 in Private Law (Université Paris Ouest), a lawyer certificate (CAPA) and a Diploma In English Law (University of Westminster - London).
In 2007, she is the head of the legal department at SAGEM WIRELESS.
Lawyer in the Paris Chamber, she is specialised in industrials and commercials agreements in France and worldwide.
Today, she writes contrats and assists companies in their protection strategy and intellectual property.
She teaches Business Law at CY Université.

At first, personal data were not protected. It started by the UN in 1948 as everyone has a right of privacy like home protection, medical and profesional secrecy and privacy.
The GDPR was created in 2018 following the CNIL creation in 1978.
The objective is to standardise data protection laws in the EU while processing personal data.
This reform aims to consolidate the rights of individuals and to introduce the accountability of responsabilities.
Every organisations are concerned by GDPR.
It includes personal data and sensitive data.
Personal data are about a natural person that can be identified directly or indirectly (identity, personal life, career, diploma, GPS, taxes).
Sensitive data are the disclosure of sensitive information (health, sexual orientation).
Processing sensitve data is prohibited except if you have given your consent or if it is vital.
Therefore, information only become data if it is recorded.
Now data collectors must prove that they are compliant with GDPR and respect privacy.

He is the scientist director of Onepoint, consulting company.
Theoritical physicist, he is specialised in AI, Machine Learning and NLP.
He is the co-author of "The automatic processing of languages" published in february 2020.

NLP is a computational linguistics.
NLP is used in automatic translation and spell checking, processing and analysing corpus document, simulation conversation (chat box), virtual assistants, automatic text generation (statistics used for making text understantable for all), automatic text generation, document classification, text summarisation, matching text with a set of criteria.
We use NLP in health care (analysing a vast amount of publications), in public opinions with surveys.
The elementary tasks for NLP are : language detection, summarising, speech tagging, question answering, syntax analysis, translation, terminological extraction.
A remarkable fact is that a significant part of the syntax and sementics of a language can be learned throught a purely statistical approach, without using any explicit rules or abstract symbolic representations.
You can solve sementic ambiguities using the attention mechanism.

The criteria for choosing a generic task for transfert learning are :

1/ the task to be learned should require solving hard problems of NLP : such as the mastery of syntax, sementics, logical deduction. In case the model can solve it, this will mean it "understands" language.

2/ a task for which training data is already available in quantity and cost free.

Several years ago, scientists used to perform science and the grants they obtained were based on merit. Their results were then used by policy makers to inform health policy.

Nowadays, scientists still perform science, but sometimes industry provides funding to scientists, making them loose their independency and making science too complicated for policy makers to interpret.

Old medicine used to deal with infections and microbes while now it has to deal with chronicle diseases and multinational corporations.

A striking example is given with the consumption of sugar. Many scientists had their researches about the danger of a high consumption of sugar funded by companies like Coca Cola. Therefore, some scientific papers claimed that obesity and diabetes were not due to sugar but to a lack of physical activity.

Scientists like Robert Lusting fought against those theories. Rober Lustig is the godfather of the sugar tax. He showed that in order to reach their goal, scientists have to: advocate themselves, advocate their position and advocate for science.

One mean to fight against the dark force is to write Op-Eds in order to influence opinion. In these Op-Eds, you must demonstrate your lack of bias, by being up front with your values otherwise people will not trust you.

• You must be persuasive and short

• You must use data, but not too much

• You must teach them something they did not know

• You must be science driven— the science is your “sword” and “shield”

So you have to (quickly):

• delineate the problem
• delineate your interest in the problem
• delineate the science behind the problem
• delineate the solution in general terms

In the United States, rates of suicide, depression and death from heroin have grown up severely. Deaths due to diabetes or heart diseases are also increasing seriously but the money is going to chronic metabolic diseases and not to hospitals.
Mental health, metabolic health and the social security system are going down to the tubes. Those problems are related. Is one of them the cause of the others or is there another factor?

Data show that there is a link between happiness and longevity. However, we have to distinguish pleasure from happiness as they are completely different:

• Pleasure is characterised by: short lived, visceral, taking, experienced alone, achievable with substances, extremes lead to addictions, dopamine.
• Happiness is related to: long lived, giving, experienced with others, not achievable with substances, cannot be addicted to happiness, serotonin.

The Prefrontal Cortex (PFC) is the brain’s brake to addiction, depression, anxiety, inattention, hate. The PFC thickness is related to obesity and executive functions and mainly due to too high dopamine and cortisol levels and low level of serotonin.
In the United States, Data show that in the same states, we have obesity, less active people, unhappy people, high diabetes, heart and Alzheimer’s rates, high opioid and high soda rates per capita.
Two main factors increase the level of dopamine: sugar and addiction to social network as Facebook. Data show that the behaviors due to high level of dopamine increased with the arrival of processed food and IPhone.

Doctorat Sorbonne Paris Cité ‘Eigenvalues and eigenvectors of large matrices under random perturbation’
Technology Director AgenT Biotech Paris Station F
Co-founder of Think Tank 4th revolution

Machine learning and deep learning, two major sub-domains of artificial intelligence, are more and more used in various fields especially in healthcare. Indeed, over the past decade, impressive progress have been accomplished in medicine thanks to the capacity of artificial intelligence to analyze complex biological data by taking into account a very large number of parameters.

As examples, we can mention:

- In cardiology: The electrocardiogram is a test as simple to perform as it is difficult to interpret. The French startup Cardiologs has set up an artificial intelligence algorithm allowing to detect in only a few minutes and only on the basis of an electrocardiogram several cardiac pathologies, something that very often requires extensive efforts from cardiologists.

- In radiology: Many newspaper articles enthusiastically show the progress of AI in the early detection of tumours on X-rays, in particular by showing how well AI is outperforming radiologists. However, although machines outperform physicians in this particular area, it is interesting to note that humans and machines do not make the same type of mistakes. Thus, the best detection algorithm is one that combines the predictions of both human and machine.

- In ophthalmology: the Google Deepmind company has set up an AI capable of diagnosing 50 eye diseases from an image of the iris. In particular, this study has revealed a previously unknown fact: this AI has succeeded in classifying Men and Women from an iris image, so the iris contains gender information.

- In diagnostics: Electronic noses (or e-noses) consisting of miniaturized arrays of broad-spectrum chemical sensors, coupled with signal processing by AI are capable of diagnosing several diseases such as cancer or diabetes with great accuracy from the patient's breath. This study has thus confirmed an old intuition: diseases have specific odors.

Nevertheless, several major difficulties and challenges are still present in the world of AI applied to medicine.
We can mention the quasi-systematic non-linearity of biological data, the so-called cohort effects, the required high dimensionality (ie sometimes a very large quantity of biological parameters must be considered to predict a disease), the uncertainty of patient annotations due to diagnostic errors or medical follow-up not long enough...

And one of the most important matters is that of the explicability of the decisions made by the AI. Indeed, it is necessary to be able to justify all the medical decisions taken by these algorithms by concerns of responsibility but also by concerns of advancing research through a better understanding of the biological functions studied. This is why, contrary to other fields of AI applications, here the pure deeplearning algorithms that can be described as black boxes, however powerful they may be, must often be replaced by algorithms that are more explicable and built thanks to the enlightened collaboration of physicians and mathematicians.

In this presentation, we discussed some of the solutions to these problems, including algorithms that can describe the importance and synergy of different biological parameters for a given disease, as well as transfer learning techniques that allow us to understand complex biological problems by transferring knowledge gained from studying other problems that are similar but simpler than the first.

Finally, through the new capacity offered by AI to analyze complex biological mechanisms requiring the consideration of a very large number of parameters, we can expect from tomorrow's medicine a better understanding of sciences such as metabolomics or genomics. These latter fields are already leading us towards diagnostics and therapies that are increasingly personalized and consequently effective, and we can therefore expect today's medicine, which is essentially curative, to be transformed tomorrow into an essentially preventive, or even corrective medicine.

Charlotte BARBIER

PHDs in Educational Sciences at Université de Paris, working on the way teachers concive critical thinking
Master Research in Educational Sciences
Bachelor on Language Didactic, Educational Sciences and Korean


Why is critical thinking important ?

‘Critical thinking is a reasonable and reflective thinking focused on deciding what to believe or do’.


There is two understandings of the notion :

1/ goal of science education, link with scientific approach while distancing yourself from your own conceptions, how do we know what we know ?

2/ link with social concerns, you need to develop your curiosity, question, be flexible, open and show humility.


Thinking criticaly is not easy !


How can it relate to media literacy ?

Media literacy is the ability to access, analyse, evaluate and create media.

We spend a huge amount of time consuming media : how can we judge the reliability of an information ?

‘True neutrality does not exist but objectivity does’.

‘An opinion is not a fact’.


How can we apply critical thinking to ourselves ?

Alexandre Pasche

Master in Political Sciences in Nanterre University
Worked for 10 years in advertisement (PUBLICIS, HAVAS)
2003 : creates is own ad agency ‘ECO and CO’ specialised in ecology and sustainable development
Communication Professor in CY Cergy Paris Université
SDSN (Sustainable Development Solutions Network) France delegate

Life is cool in 2020 : technology makes life easier. We have acces to every music and culture for free.

We have faced 50 years of great progress : life expectancy has increased by 8 years, literacy rates are higher, the air is less polluted in developed countries.

But there is a negative side of the coin : cell phones batteries pollute water, some countries are disappearing like 80% of insects, only 6% of the plastic produced since 1980 has been recycled.

We are now facing unbelievable inequalities with consequences for all (air pollution, dangerous products effects).

We must invent a new kind of developement : a sustainable development.

Alonso Silva

Currently working at Safran as Senior Machine Learning researcher
2012/2018 : Bell Labs
2011/2012 : postdoctoral researcher in the Department of Electrical Engineering and Computer Sciences at the University of California, Berkeley
2010/2011 : postdoctoral researcher in the Department of INRIA Paris Rocquencourt
2010 : Ph.D. in Physics from the École Supérieure d'Électricité Ph.D. at INRIA Sophia-Antipolis, France
2006 : B.Sc. of Mathematical Engineering and Mathematical Engineering degree from the Department of Mathematical Engineering (DIM) at the Universidad de Chile

Every week, our students assist to a lecture which is in link with the current world and the Bachelor programme.

This enlightment session explained to students what survival analysis is.

It was historically developed and used by actuaries in insurance and in medical researches to measure lifetime of population and expected lifetime of patients.

Alonso applied this concept to several modern examples of survival analyses using the Kaplan Meiser estimation like the duration of a dictatorship versus a democracy or analysing the customer churn data using a logistic regression.

"Survival analysis is useful when your data has a bith, a death and a right censorship". Alonso uses this concept to estimate the life expectation of planes and helicopters of the Safran fleets.

"Machine Learning can help us to better understand datas".