We consider a transfer-learning problem by using the parameter transfer approach, where a suitable parameter of feature mapping is learned through one task and applied to another objective task. Then, we introduce the notion of the local stability of parametric feature mapping and parameter transfer learnability, and thereby derive a learning bound for parameter transfer algorithms. As an application of parameter transfer learning, we discuss the performance of sparse coding in self-taught learning. Although self-taught learning algorithms with plentiful unlabeled data often show excellent empirical performance, their theoretical analysis has not been studied. In this paper, we also provide the first theoretical learning bound for self-taught learning.

The training and evaluation scripts support the following problems (see util.py for more details): New problems can be implemented very easily. You can see in train.py Disclaimer: This is not an official Google product.

Why is it, when we need to learn something new that will benefit our work or home life, we often find ourselves blocked by seemingly invisible forces? When learning fails we miss out on important projects, promotions, and opportunities. We end up suffering and falling short of our objectives. Somehow, for many of us, our natural capacity to learn seemed to deteriorate over time, especially in areas that we care about the most. Educators and business leaders have used the term "learning to learn" to name a missing skill that would reverse the deterioration.

Due to physiological variation, patients diagnosed with the same condition may exhibit divergent, but related, responses to the same treatments. Hidden Parameter Markov Decision Processes (HiP-MDPs) tackle this transfer-learning problem by embedding these tasks into a low-dimensional space. However, the original formulation of HiP-MDP had a critical flaw: the embedding uncertainty was modelled independently of the agent's state uncertainty, requiring an unnatural training procedure in which all tasks visited every part of the state space--possible for robots that can be moved to a particular location, impossible for human patients. We update the HiP-MDP framework and extend it to more robustly develop personalized medicine strategies for HIV treatment.

We aim to create a framework for transfer learning using latent factor models to learn the dependence structure between a larger source dataset and a target dataset. The methodology is motivated by our goal of building a risk-assessment model for surgery patients, using both institutional and national surgical outcomes data. The national surgical outcomes data is collected through NSQIP (National Surgery Quality Improvement Program), a database housing almost 4 million patients from over 700 different hospitals. We build a latent factor model with a hierarchical prior on the loadings matrix to appropriately account for the different covariance structure in our data. We extend this model to handle more complex relationships between the populations by deriving a scale mixture formulation using stick-breaking properties. Our model provides a transfer learning framework that utilizes all information from both the source and target data, while modeling the underlying inherent differences between them.

The meetup will be hosted by AAA-data / Comité des Constructeurs Français d'Automobiles and the networking event is sponsored by Zen.ly . A big thank you to them. The program for this third regular meetup of the season (and the fifth total for season 4) is a little extraordinary this time and will feature the following: Dominique Gillot, Sénatrice, ancienne ministre et Rapporteure avec le député Claude de Ganay d'un rapport sur l'Intelligence Artificielle pour le Parlement. Julien Simon (Amazon), Machine Learning & Deep Learning with Amazon Web Services Gautier Marti (Hellebore Capital), A closer look at correlations You may have already read many times that the job of a Data Scientist is to skim through a huge amount of data searching for correlations between some variables of interest. And also, that one of his worst enemies (besides correlation doesn't imply causation) is spurious correlation.

Communities of people are often modelled as social networks consisting of individual actors whose roles in the community correspond to the network patterns present around their corresponding nodes. Examples of such roles for individual actors in social networks are people bridging two communities, central people through which a large part of communication passes, and outliers. In social network analysis, recognising user roles is helpful to gain deeper understanding of the underlying communities. For large online social networks, the only scalable way to achieve this is through automatic labelling of nodes, i.e. using machine learning. If, in a community, persons are already annotated with roles (by whatever method), this can be exploited to train a classifier to detect person roles in case new people appear in the community.

We consider the problem of transfer learning in an online setting. Different tasks are presented sequentially and processed by a within-task algorithm. We propose a lifelong learning strategy which refines the underlying data representation used by the within-task algorithm, thereby transferring information from one task to the next. We show that when the within-task algorithm comes with some regret bound, our strategy inherits this good property. Our bounds are in expectation for a general loss function, and uniform for a convex loss. We discuss applications to dictionary learning and finite set of predictors. In the latter case, we improve previous $O(1/\sqrt{m})$ bounds to $O(1/m)$ where $m$ is the per task sample size.

Abstract--The ability of a human being to extrapolate previously gained knowledge to other domains inspired a new family of methods in machine learning called transfer learning. Transfer learning is often based on the assumption that objects in both target and source domains share some common feature and/or data space. In this paper, we propose a simple and intuitive approach that minimizes iteratively the distance between source and target task distributions by optimizing the kernel target alignment (KT A). We show that this procedure is suitable for transfer learning by relating it to Hilbert-Schmidt Independence Criterion (HSIC) and Quadratic Mutual Information (QMI) maximization. We run our method on benchmark computer vision data sets and show that it can outperform some state-of-art methods. I NTRODUCTION Most research in machine learning is usually concentrated around the setting where a classifier is trained and tested on data drawn from the same distribution. This scenario has already been well investigated and in some tasks supervised approaches have almost no room for improvement. However, building humanlike intelligent systems requires them to be able to generalize the discovered patterns to previously unseen domains.

In this paper, we propose a novel learning framework for the problem of domain transfer learning. We map the data of two domains to one single common space, and learn a classifier in this common space. Then we adapt the common classifier to the two domains by adding two adaptive functions to it respectively. In the common space, the target domain data points are weighted and matched to the target domain in term of distributions. The weighting terms of source domain data points and the target domain classification responses are also regularized by the local reconstruction coefficients. The novel transfer learning framework is evaluated over some benchmark cross-domain data sets, and it outperforms the existing state-of-the-art transfer learning methods.