"In a world more and more driven by AI models, Data Scientists cannot effectively ascertain on their own the costs associated with the unintended consequences of False Positives and False Negatives. Mitigating unintended consequences requires the collaboration across a diverse set of stakeholders in order to identify the metrics against which the AI Utility Function will seek to optimize." I've been fortunate enough to have had some interesting conversations since publishing that blog, especially with an organization who is championing data ethics and "Responsible AI" (love that term). As was so well covered in Cathy O'Neil's book "Weapons of Math Destruction", the biases built into many of the AI models that are being used to approve loans and mortgages, hire job applicants, and accept university admissions are yielding unintended consequences that severely impact both individuals and society. AI models only optimize against the metrics against which it has been programmed to optimize.

My statistics course is ideal for those studying on their own, or if you are in a statistics class and struggling with your assigned textbook or lecture material. I know stats courses can be boring, so I try to make it as exciting as possible. The examples have a psychology bend, but this course is absolutely relevant for business students, especially those in data analysis that have to get a better understanding of statistical tests and the fundamental concepts. So welcome to you whether you are in business or psychology. I provide examples within the lessons, so this should cut down your study time. Furthermore, I make sure that students understand the links between the different lessons.

I recently started to follow an exciting and mind-bending philosophy online course at MIT called Minds and Machines. The course is a thorough, rigorous 12 Weeks Learning Path introduction to contemporary philosophy of mind, exploring consciousness, reality, artificial intelligence (AI), and more. It is definitively one of the most in-depth philosophy courses available online that I ever frequented. The first effect of starting study philosophy at Massachusetts Institute of Technology is that I'm asking more challenging questions… the second effect is that I'm writing more about those questions. I'm in this moment, exploring the relationship between the mind and the body, the capacity of computers to think, the way we perceive reality, and the perspective of the existence of a science of consciousness. As a first result, I've started to pay particular attention to one specific question that definitively has a lot to relate to my daily work as an AI expert: what is intelligence?

Ethics will be at the forefront of robotics education thanks to a new University of Texas at Austin program that will train tomorrow's technologists to understand the positive -- and potentially negative -- implications of their creations. Today, much robotic technology is developed without considering its potentially harmful effects on society, including how these technologies can infringe on privacy or further economic inequity. The new UT Austin program will fill an important educational gap by prioritizing these issues in its curriculum. "In the next 10 years, we are going to live more closely alongside robots, and we want to be sure that those robots are fair, inclusive and free from bias," said Junfeng Jiao, associate professor in the School of Architecture and the program lead. "And because the robots we create are reflections of ourselves, it is imperative that technologists receive an excellent ethics education. We want our students to work directly with companies to create practices and technologies that are equitable and fair."

Search engines answer users' queries by listing relevant items (e.g. documents, songs, products, web pages, ...). These engines rely on algorithms that learn to rank items so as to present an ordered list maximizing the probability that it contains relevant item. The main challenge in the design of learning-to-rank algorithms stems from the fact that queries often have different meanings for different users. In absence of any contextual information about the query, one often has to adhere to the {\it diversity} principle, i.e., to return a list covering the various possible topics or meanings of the query. To formalize this learning-to-rank problem, we propose a natural model where (i) items are categorized into topics, (ii) users find items relevant only if they match the topic of their query, and (iii) the engine is not aware of the topic of an arriving query, nor of the frequency at which queries related to various topics arrive, nor of the topic-dependent click-through-rates of the items. For this problem, we devise LDR (Learning Diverse Rankings), an algorithm that efficiently learns the optimal list based on users' feedback only. We show that after $T$ queries, the regret of LDR scales as $O((N-L)\log(T))$ where $N$ is the number of all items. We further establish that this scaling cannot be improved, i.e., LDR is order optimal. Finally, using numerical experiments on both artificial and real-world data, we illustrate the superiority of LDR compared to existing learning-to-rank algorithms.

Constraint acquisition systems such as QuAcq and MultiAcq can assist non-expert users to model their problems as constraint networks by classifying (partial) examples as positive or negative. For each negative example, the former focuses on one constraint of the target network, while the latter can learn a maximum number of constraints. Two bottlenecks of the acquisition process where both these algorithms encounter problems are the large number of queries required to reach convergence, and the high cpu times needed to generate queries, especially near convergence. In this paper we propose algorithmic and heuristic methods to deal with both these issues. We first describe an algorithm, called MQuAcq, that blends the main idea of MultiAcq into QuAcq resulting in a method that learns as many constraints as MultiAcq does after a negative example, but with a lower complexity. A detailed theoretical analysis of the proposed algorithm is also presented. %We also present a technique that boosts the performance of constraint acquisition by reducing the number of queries significantly. Then we turn our attention to query generation which is a significant but rather overlooked part of the acquisition process. We describe %in detail how query generation in a typical constraint acquisition system operates, and we propose heuristics for improving its efficiency. Experiments from various domains demonstrate that our resulting algorithm that integrates all the new techniques does not only generate considerably fewer queries than QuAcq and MultiAcq, but it is also by far faster than both of them, in average query generation time as well as in total run time, and also largely alleviates the premature convergence problem.

Automatically discovering a process model from an event log is the prime problem in process mining. This task is so far approached as an unsupervised learning problem through graph synthesis algorithms. Algorithmic design decisions and heuristics allow for efficiently finding models in a reduced search space. However, design decisions and heuristics are derived from assumptions about how a given behavioral description - an event log - translates into a process model and were not learned from actual models which introduce biases in the solutions. In this paper, we explore the problem of supervised learning of a process discovery technique D. We introduce a technique for training an ML-based model D using graph convolutional neural networks; D translates a given input event log into a sound Petri net. We show that training D on synthetically generated pairs of input logs and output models allows D to translate previously unseen synthetic and several real-life event logs into sound, arbitrarily structured models of comparable accuracy and simplicity as existing state of the art techniques for discovering imperative process models. We analyze the limitations of the proposed technique and outline alleys for future work.

Counterfactual explanations focus on "actionable knowledge" to help end-users understand how a machine learning outcome could be changed to a more desirable outcome. For this purpose a counterfactual explainer needs to discover input dependencies that relate to outcome changes. Identifying the minimum subset of feature changes needed to action an output change in the decision is an interesting challenge for counterfactual explainers. The DisCERN algorithm introduced in this paper is a case-based counter-factual explainer. Here counterfactuals are formed by replacing feature values from a nearest unlike neighbour (NUN) until an actionable change is observed. We show how widely adopted feature relevance-based explainers (i.e. LIME, SHAP), can inform DisCERN to identify the minimum subset of "actionable features". We demonstrate our DisCERN algorithm on five datasets in a comparative study with the widely used optimisation-based counterfactual approach DiCE. Our results demonstrate that DisCERN is an effective strategy to minimise actionable changes necessary to create good counterfactual explanations.

Disclaimer: This is my personal Medium blog, therefore anything I post, share, and comment don't reflect my employer. When we started exploring Machine Learning, we realized that there were many solutions available in the market from concept to production. To simplify and understand the options, our initial approach has been, understanding of concepts from MOOC's, hands on coding using classical ML packages like Numpy, Pandas, Scikit Learn, NLTK, Matplotlib…mostly using Python/R on Anaconda/Pycharm/Rstudio as IDE Though it was easy to get hands dirty & see the power of ML, it became complex when the classical ML approach were deployed for real world problems. Problems compounded due to many factors such as availability /quality of training data, imbalanced data sets, feature engineering requirements, lower precision, accuracy…. When we were building a Machine learning-based classifier for a NLP Classification problem for a business requirement, we encountered challenges in the form of Feature Selection, Imbalanced Data Set, hyper parameter Tuning, etc.

As a student, who has been in school almost my entire life, I always felt that the knowledge we learn in school is theoretical. That feeling only grew whenever I talked to my friends and colleagues who worked in the industry or had some experiences outside the university. This gap between what we learn in our degrees and what we actually need to succeed and build a career is one of the reasons internships are necessary for any student. Although internships are important for basically any student regardless of their major, it's even more important if you're a student in a tech field. In tech, we study the history of a field, the tools we can use, and some field applications.