This paper describes the SemEval-2020 shared task "Assessing Humor in Edited News Headlines." The task's dataset contains news headlines in which short edits were applied to make them funny, and the funniness of these edited headlines was rated using crowdsourcing. This task includes two subtasks, the first of which is to estimate the funniness of headlines on a humor scale in the interval 0-3. The second subtask is to predict, for a pair of edited versions of the same original headline, which is the funnier version. To date, this task is the most popular shared computational humor task, attracting 48 teams for the first subtask and 31 teams for the second.

Federated learning has received significant interests recently due to its capability of learning a shared machine learning model across smart devices without accessing their private data in the era of Internet of things. This paper jointly considers two critical issues of federated learning -- data privacy and communication efficiency -- and develops a communication-efficient and differentially-private federated learning scheme called CPFed. The main challenge in addressing both issues together lies in the fact that data compression techniques often lead to an increased number of training iterations required for achieving some desired training loss due to the compression errors, while the differential privacy guarantee usually deteriorates with respect to the number of training iterations. To reconcile this dilemma, we propose to use sparsified privacy-masking that first adds random noise to the model update and then applies unbiased random sparsifier before uploading the model update at each device in federated learning. By using sparsified privacy-masking, our proposed CPFed scheme can achieve high communication efficiency and strong data privacy guarantee at the same time while preserving model accuracy. We provide an explict end-to-end privacy guarantee of CPFed using zero-concentrated differential privacy and give its theoretical convergence rates for both convex and non-convex models. Through extensive numerical experiments on real-world datasets, we demonstrate the effectiveness and efficiency of our proposed method.

The performance of many machine learning models depends on their hyper-parameter settings. Bayesian Optimization has become a successful tool for hyper-parameter optimization of machine learning algorithms, which aims to identify optimal hyper-parameters during an iterative sequential process. However, most of the Bayesian Optimization algorithms are designed to select models for effectiveness only and ignore the important issue of model training efficiency. Given that both model effectiveness and training time are important for real-world applications, models selected for effectiveness may not meet the strict training time requirements necessary to deploy in a production environment. In this work, we present a unified Bayesian Optimization framework for jointly optimizing models for both prediction effectiveness and training efficiency. We propose an objective that captures the tradeoff between these two metrics and demonstrate how we can jointly optimize them in a principled Bayesian Optimization framework. Experiments on model selection for recommendation tasks indicate models selected this way significantly improves model training efficiency while maintaining strong effectiveness as compared to state-of-the-art Bayesian Optimization algorithms.

We initiate the study of a new model of supervised learning under privacy constraints. Imagine a medical study where a dataset is sampled from a population of both healthy and unhealthy individuals. Suppose healthy individuals have no privacy concerns (in such case, we call their data "public") while the unhealthy individuals desire stringent privacy protection for their data. In this example, the population (data distribution) is a mixture of private (unhealthy) and public (healthy) sub-populations that could be very different. Inspired by the above example, we consider a model in which the population $\mathcal{D}$ is a mixture of two sub-populations: a private sub-population $\mathcal{D}_{\sf priv}$ of private and sensitive data, and a public sub-population $\mathcal{D}_{\sf pub}$ of data with no privacy concerns. Each example drawn from $\mathcal{D}$ is assumed to contain a privacy-status bit that indicates whether the example is private or public. The goal is to design a learning algorithm that satisfies differential privacy only with respect to the private examples. Prior works in this context assumed a homogeneous population where private and public data arise from the same distribution, and in particular designed solutions which exploit this assumption. We demonstrate how to circumvent this assumption by considering, as a case study, the problem of learning linear classifiers in $\mathbb{R}^d$. We show that in the case where the privacy status is correlated with the target label (as in the above example), linear classifiers in $\mathbb{R}^d$ can be learned, in the agnostic as well as the realizable setting, with sample complexity which is comparable to that of the classical (non-private) PAC-learning. It is known that this task is impossible if all the data is considered private.

The Uniform Manifold Approximation and Projection (UMAP) algorithm has become widely popular for its ease of use, quality of results, and support for exploratory, unsupervised, supervised, and semi-supervised learning. While many algorithms can be ported to a GPU in a simple and direct fashion, such efforts have resulted in inefficent and inaccurate versions of UMAP. We show a number of techniques that can be used to make a faster and more faithful GPU version of UMAP, and obtain speedups of up to 100x in practice. Many of these design choices/lessons are general purpose and may inform the conversion of other graph and manifold learning algorithms to use GPUs. Our implementation has been made publicly available as part of the open source RAPIDS cuML library (https://github.com/rapidsai/cuml).

Gaussian processes are distributions over functions that are versatile and mathematically convenient priors in Bayesian modelling. However, their use is often impeded for data with large numbers of observations, $N$, due to the cubic (in $N$) cost of matrix operations used in exact inference. Many solutions have been proposed that rely on $M \ll N$ inducing variables to form an approximation at a cost of $\mathcal{O}(NM^2)$. While the computational cost appears linear in $N$, the true complexity depends on how $M$ must scale with $N$ to ensure a certain quality of the approximation. In this work, we investigate upper and lower bounds on how $M$ needs to grow with $N$ to ensure high quality approximations. We show that we can make the KL-divergence between the approximate model and the exact posterior arbitrarily small for a Gaussian-noise regression model with $M\ll N$. Specifically, for the popular squared exponential kernel and $D$-dimensional Gaussian distributed covariates, $M=\mathcal{O}((\log N)^D)$ suffice and a method with an overall computational cost of $\mathcal{O}(N(\log N)^{2D}(\log\log N)^2)$ can be used to perform inference.

Summary: Building classification models that predict a binary class label on the basis of high dimensional multi-omics datasets poses several challenges, due to the typically widely differing characteristics of the data layers in terms of number of predictors, type of data, and levels of noise. Previous research has shown that applying classical logistic regression with elastic-net penalty to these datasets can lead to poor results (Liu et al., 2018). We implement a two-step approach to multi-omic logistic regression in which variable selection is performed on each layer separately and a predictive model is then built using the variables selected in the first step. Here, our approach is compared to other methods that have been developed for the same purpose, and we adapt existing software for multi-omic linear regression (Zhao and Zucknick, 2020) to the logistic regression setting. Extensive simulation studies show that our approach should be preferred if the goal is to select as many relevant predictors as possible, as well as achieving prediction performances comparable to those of the best competitors. Our motivating example is a cardiometabolic syndrome dataset comprising eight'omic data types for 2 extreme phenotype groups (10 obese and 10 lipodystrophy individuals) and 185 blood donors. Our proposed approach allows us to identify features that characterise cardiometabolic syndrome at the molecular level.

This paper explores the problem of multi-view spectral clustering (MVSC) based on tensor low-rank modeling. Unlike the existing methods that all adopt an off-the-shelf tensor low-rank norm without considering the special characteristics of the tensor in MVSC, we design a novel structured tensor low-rank norm tailored to MVSC. Specifically, we explicitly impose a symmetric low-rank constraint and a structured sparse low-rank constraint on the frontal and horizontal slices of the tensor to characterize the intra-view and inter-view relationships, respectively. Moreover, the two constraints could be jointly optimized to achieve mutual refinement. On the basis of the novel tensor low-rank norm, we formulate MVSC as a convex low-rank tensor recovery problem, which is then efficiently solved with an augmented Lagrange multiplier based method iteratively. Extensive experimental results on five benchmark datasets show that the proposed method outperforms state-of-the-art methods to a significant extent. Impressively, our method is able to produce perfect clustering. In addition, the parameters of our method can be easily tuned, and the proposed model is robust to different datasets, demonstrating its potential in practice.

I think that the best way to really understand how a neural network works is to implement one from scratch. That is exactly what I going to do through this article. I will create a neural network class, and I want to design it in such a way to be more flexible. I do not want to hardcode in it a specific activation or loss functions, or optimizers (that is SGD, Adam, or other gradient-based methods). I will design it to receive these from outside the class so that one can just take the class's code and pass to it whatever activation/loss/optimizer he wants.

Reading through a data science book or taking a course, it can feel like you have the individual pieces, but don't quite know how to put them together. Taking the next step and solving a complete machine learning problem can be daunting, but preserving and completing a first project will give you the confidence to tackle any data science problem. This series of articles will walk through a complete machine learning solution with a real-world dataset to let you see how all the pieces come together. We'll follow the general machine learning workflow step-by-step: Along the way, we'll see how each step flows into the next and how to specifically implement each part in Python. The complete project is available on GitHub, with the first notebook here. After completing the work, I was offered the job, but then the CTO of the company quit and they weren't able to bring on any new employees. I guess that's how things go on the start-up scene!) The first step before we get coding is to understand the problem we are trying to solve and the available data. In this project, we will work with publicly available building energy data from New York City. The objective is to use the energy data to build a model that can predict the Energy Star Score of a building and interpret the results to find the factors which influence the score. We want to develop a model that is both accurate *-- it can predict the Energy Star Score close to the true value -- and *interpretable -- we can understand the model predictions. Once we know the goal, we can use it to guide our decisions as we dig into the data and build models. Contrary to what most data science courses would have you believe, not every dataset is a perfectly curated group of observations with no missing values or anomalies (looking at you mtcars and iris datasets). Real-world data is messy which means we need to clean and wrangle it into an acceptable format before we can even start the analysis. Data cleaning is an un-glamorous, but necessary part of most actual data science problems.