Understanding the structure of multiple related tasks allows for multi-task learning to improve the generalisation ability of one or all of them. However, it usually requires training each pairwise combination of tasks together in order to capture task relationships, at an extremely high computational cost. In this work, we learn task relationships via an automated weighting framework, named Auto-Lambda. Unlike previous methods where task relationships are assumed to be fixed, Auto-Lambda is a gradient-based meta learning framework which explores continuous, dynamic task relationships via task-specific weightings, and can optimise any choice of combination of tasks through the formulation of a meta-loss; where the validation loss automatically influences task weightings throughout training. We apply the proposed framework to both multi-task and auxiliary learning problems in computer vision and robotics, and show that Auto-Lambda achieves state-of-the-art performance, even when compared to optimisation strategies designed specifically for each problem and data domain. Finally, we observe that Auto-Lambda can discover interesting learning behaviors, leading to new insights in multi-task learning. Code is available at https://github.com/lorenmt/auto-lambda.

Pre-trained protein models (PTPMs) represent a protein with one fixed embedding and thus are not capable for diverse tasks. For example, protein structures can shift, namely protein folding, between several conformations in various biological processes. To enable PTPMs to produce task-aware representations, we propose to learn interpretable, pluggable and extensible protein prompts as a way of injecting task-related knowledge into PTPMs. In this regard, prior PTPM optimization with the masked language modeling task can be interpreted as learning a sequence prompt (Seq prompt) that enables PTPMs to capture the sequential dependency between amino acids. To incorporate conformational knowledge to PTPMs, we propose an interaction-conformation prompt (IC prompt) that is learned through back-propagation with the protein-protein interaction task. As an instantiation, we present a conformation-aware pre-trained protein model that learns both sequence and interaction-conformation prompts in a multi-task setting. We conduct comprehensive experiments on nine protein datasets. Results confirm our expectation that using the sequence prompt does not hurt PTPMs' performance on sequence-related tasks while incorporating the interaction-conformation prompt significantly improves PTPMs' performance on tasks where conformational knowledge counts. We also show the learned prompts can be combined and extended to deal with new complex tasks.

In practice, data augmentation is assigned a predefined budget in terms of newly created samples per epoch. When using several types of data augmentation, the budget is usually uniformly distributed over the set of augmentations but one can wonder if this budget should not be allocated to each type in a more efficient way. This paper leverages online learning to allocate on the fly this budget as part of neural network training. This meta-algorithm can be run at almost no extra cost as it exploits gradient based signals to determine which type of data augmentation should be preferred. Experiments suggest that this strategy can save computation time and thus goes in the way of greener machine learning practices.

Principal component analysis is a simple yet useful dimensionality reduction technique in modern machine learning pipelines. In consequential domains such as college admission, healthcare and credit approval, it is imperative to take into account emerging criteria such as the fairness and the robustness of the learned projection. In this paper, we propose a distributionally robust optimization problem for principal component analysis which internalizes a fairness criterion in the objective function. The learned projection thus balances the trade-off between the total reconstruction error and the reconstruction error gap between subgroups, taken in the min-max sense over all distributions in a moment-based ambiguity set. The resulting optimization problem over the Stiefel manifold can be efficiently solved by a Riemannian subgradient descent algorithm with a sub-linear convergence rate. Our experimental results on real-world datasets show the merits of our proposed method over state-of-the-art baselines. Machine learning models are ubiquitous in our daily lives and supporting the decision-making process in diverse domains.

The artificial intelligence (AI) "strategic competition" with China is more intense than ever. To many, the stakes have never been higher -- who leads in AI will lead globally. At first glance, China appears to be well-positioned to take the lead when it comes to AI talent. China is actively integrating AI into every level of its education system, while the United States has yet to embrace AI education as a strategic priority. To maintain its competitive edge, the United States must adopt AI education and workforce policies that are targeted and coordinated.

A free online version of the second edition of the book based on Stat 110, Introduction to Probability by Joe Blitzstein and Jessica Hwang, is now available here. Print copies are available via CRC Press, Amazon, and elsewhere. Stat110x is also available as an free edX course, here. The edX course focuses on animations, interactive features, readings, and problem-solving, and is complementary to the Stat 110 lecture videos on YouTube, which are available here. The Stat110x animations are available within the course and here.

You have finished solving a machine learning problem. The accuracy of your model is awesome. Till now your work is probably a Jupyter notebook, which is full of code, a few visuals, and very little documentation. If you see your work after a month or so, you might struggle to understand your own creation. To make matter worse, the Jupyter notebook does not have all decisions and assumptions you have taken in the solution.

Over the past few years, we have seen a massive advancement in technology. Every time a new tech is born, we see it getting applied to various industries. Artificial intelligence is the most popular technology of this decade, and it is helping in solving some of society's most challenging adversities and provides a safer, healthier and more prosperous world for all. We've already shared some of the exciting possibilities in the fields of Education, Businesses in our blog section. But there may be no field where the chances are more exciting than education and skills. In the last blog, we have seen how AI is helping in learning.

We are a team of Professionals having domain expertise in various fields. All the team members are highly committed and enthusiastic about education. Our team members possess hands on experience in teaching to a varied age groups starting age 6 years to working professionals. We know, Education needs to be practical and all of our courses are based on our practical experience,we try to explain even theoretical concepts practically. We do not develop courses where we don't have practical expertise.

Knowledge of calculus is not required to get results and solve problems in machine learning or deep learning. However, knowing some calculus will help you in a number of ways, such as in reading mathematical notation in books and papers, and in understanding the terms used to describe fitting models like "gradient," and in understanding the learning dynamics of models fit via optimization such as neural networks. Calculus is a challenging topic as taught at a university level, but you don't need to know all of calculus, just a handful of terms and methods related to numerical function optimization, central to fitting algorithms like neural networks. And the best way to get a handle on calculus is from books. In this tutorial, you will discover books on calculus for machine learning.