A fundamental trait of intelligence is the ability to achieve goals in the face of novel circumstances, such as making decisions from new action choices. However, standard reinforcement learning assumes a fixed set of actions and requires expensive retraining when given a new action set. To make learning agents more adaptable, we introduce the problem of zero-shot generalization to new actions. We propose a two-stage framework where the agent first infers action representations from action information acquired separately from the task. A policy flexible to varying action sets is then trained with generalization objectives. We benchmark generalization on sequential tasks, such as selecting from an unseen tool-set to solve physical reasoning puzzles and stacking towers with novel 3D shapes. Videos and code are available at https://sites.google.com/view/action-generalization

Human head pose estimation in images has applications in many fields such as human-computer interaction or video surveillance tasks. In this work, we address this problem, defined here as the estimation of both vertical (tilt/pitch) and horizontal (pan/yaw) angles, through the use of a single Convolutional Neural Network (ConvNet) model, trying to balance precision and inference speed in order to maximize its usability in real-world applications. Our model is trained over the combination of two datasets: 'Pointing'04' (aiming at covering a wide range of poses) and 'Annotated Facial Landmarks in the Wild' (in order to improve robustness of our model for its use on real-world images). Three different partitions of the combined dataset are defined and used for training, validation and testing purposes. As a result of this work, we have obtained a trained ConvNet model, coined RealHePoNet, that given a low-resolution grayscale input image, and without the need of using facial landmarks, is able to estimate with low error both tilt and pan angles (~4.4{\deg} average error on the test partition). Also, given its low inference time (~6 ms per head), we consider our model usable even when paired with medium-spec hardware (i.e. GTX 1060 GPU). * Code available at: https://github.com/rafabs97/headpose_final * Demo video at: https://www.youtube.com/watch?v=2UeuXh5DjAE

The paper introduces a novel framework for extracting model-agnostic human interpretable rules to explain a classifier's output. The human interpretable rule is defined as an axis-aligned hyper-cuboid containing the instance for which the classification decision has to be explained. The proposed procedure finds the largest (high \textit{coverage}) axis-aligned hyper-cuboid such that a high percentage of the instances in the hyper-cuboid have the same class label as the instance being explained (high \textit{precision}). Novel approximations to the coverage and precision measures in terms of the parameters of the hyper-cuboid are defined. They are maximized using gradient-based optimizers. The quality of the approximations is rigorously analyzed theoretically and experimentally. Heuristics for simplifying the generated explanations for achieving better interpretability and a greedy selection algorithm that combines the local explanations for creating global explanations for the model covering a large part of the instance space are also proposed. The framework is model agnostic, can be applied to any arbitrary classifier, and all types of attributes (including continuous, ordered, and unordered discrete). The wide-scale applicability of the framework is validated on a variety of synthetic and real-world datasets from different domains (tabular, text, and image).

The Travelling Salesman Problem (TSP) is a classical combinatorial optimisation problem. Deep learning has been successfully extended to meta-learning, where previous solving efforts assist in learning how to optimise future optimisation instances. In recent years, learning to optimise approaches have shown success in solving TSP problems. However, they focus on one type of TSP problem, namely ones where the points are uniformly distributed in Euclidean spaces and have issues in generalising to other embedding spaces, e.g., spherical distance spaces, and to TSP instances where the points are distributed in a non-uniform manner. An aim of learning to optimise is to train once and solve across a broad spectrum of (TSP) problems. Although supervised learning approaches have shown to achieve more optimal solutions than unsupervised approaches, they do require the generation of training data and running a solver to obtain solutions to learn from, which can be time-consuming and difficult to find reasonable solutions for harder TSP instances. Hence this paper introduces a new learning-based approach to solve a variety of different and common TSP problems that are trained on easier instances which are faster to train and are easier to obtain better solutions. We name this approach the non-Euclidean TSP network (NETSP-Net). The approach is evaluated on various TSP instances using the benchmark TSPLIB dataset and popular instance generator used in the literature. We performed extensive experiments that indicate our approach generalises across many types of instances and scales to instances that are larger than what was used during training.

Across machine learning (ML) sub-disciplines researchers make mathematical assumptions to facilitate proof-writing. While such assumptions are necessary for providing mathematical guarantees for how algorithms behave, they also necessarily limit the applicability of these algorithms to different problem settings. This practice is known--in fact, obvious-- and accepted in ML research. However, similar attention is not paid to the normative assumptions that ground this work. I argue such assumptions are equally as important, especially in areas of ML with clear social impact, such as fairness. This is because, similar to how mathematical assumptions constrain applicability, normative assumptions also limit algorithm applicability to certain problem domains. I show that, in existing papers published in top venues, once normative assumptions are clarified, it is often possible to get unclear or contradictory results. While the mathematical assumptions and results are sound, the implicit normative assumptions and accompanying normative results contraindicate using these methods in practical fairness applications.

This model finds the "best fit" line through a set of data points by using a simple formula. The variable you want to predict (the dependent variable) is represented as an equation of variables you know (independent variables). The prediction can be obtained through the outcome of the equation by inputting the independent variables, and having the equation provide the answer. The main categories of Linear models used are Linear Regression and Logistic Regression. Linear Regression is used for predicting numerical values using the "best fit" line through all data points.

This article was published as a part of the Data Science Blogathon. Every time you stumble upon a huge volume of data with thousands of features, you will be wondering what would be the best algorithm to get accurate predictions on this data, and whether to use all the features or reduce the feature space. Through this blog, I will take you through the steps in finding the good features through lasso regression and getting the right algorithm through a technique called stacking. Stacking refers to a method of joining the machine learning models, similar to arranging a stack of plates at a restaurant. It combines the output of many models.

It has been some time since we last performed a Python libraries roundup, and as such we have taken the opportunity to start the month of November with just such a fresh list. Last time we at KDnuggets did this, editor and author Dan Clark split up the vast array of Python data science related libraries up into several smaller collections, including data science libraries, machine learning libraries, and deep learning libraries. While splitting libraries into categories is inherently arbitrary, this made sense at the time of previous publication. This time, however, we have split the collected on open source Python data science libraries in two. This first post (this) covers "data science, data visualization & machine learning," and can be thought of as "traditional" data science tools covering common tasks. The second post, to be published next week, will cover libraries for use in building neural networks, and those for performing natural language processing and computer vision tasks.

The series of data points plotted against time is known as time series. It is a de-facto analysis technique used in market evaluation and in weather forecast. It is an exciting topic to study as it somehow tends to predict the future, which we are always interested in. We can make forecast of tomorrow's weather by observing the weather of past few days. If the weather was sunny for last 4–5 days then there is high chance for weather to be sunny tomorrow.

For my final capstone project in Flatiron School's Immersive Data Science Program, I decided to test my newfound skills and continue furthering my personal investigations into the relationships that exist between data, waste, and energy. Recently, I have been learning more about the various ways that Municipal Solid Waste (MSW) can be transformed into energy. The most promising and efficient technology that I have come across to date is Plasma Arc Gasification. In my research, I discovered that understanding specific composition details about the MSW to be used as feedstock is one of many critical steps in designing a plasma gasification facility. What I set out to do for my capstone project, was to see if I could find some MSW collection datasets and perform a Feedstock Analysis with the intent of calculating specific Waste Type Compositions, Energy Density (kWh/kg), and Total Energy (kWh) for each sample.