Variational Autoencoders for Collaborative Filtering

Liang, Dawen; Krishnan, Rahul G.; Hoffman, Matthew D.; Jebara, Tony

doi:10.48550/arxiv.1802.05814

Cited by 12 publications

(23 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This model may be used for artificial data generation, data representation, and inference tasks. A couple of examples are collaborative filtering (Liang et al 2018) and image analysis (Wang et al 2017).…”

Section: Autoencodersmentioning

confidence: 99%

An Information Theory Approach on Deciding Spectroscopic Follow-ups

Astudillo

Protopapas

Pichara

et al. 2019

View full text Add to dashboard Cite

Classification and characterization of variable phenomena and transient phenomena are critical for astrophysics and cosmology. These objects are commonly studied using photometric time series or spectroscopic data. Given that many ongoing and future surveys are in time-domain and given that adding spectra provide further insights but requires more observational resources, it would be valuable to know which objects should we prioritize to have spectrum in addition to time series. We propose a methodology in a probabilistic setting that determines a-priory which objects are worth taking spectrum to obtain better insights, where we focus insight as the type of the object (classification). Objects for which we query its spectrum are reclassified using their full spectrum information. We first train two classifiers, one that uses photometric data and another that uses photometric and spectroscopic data together. Then for each photometric object we estimate the probability of each possible spectrum outcome. We combine these models in various probabilistic frameworks (strategies) which are used to guide the selection of follow up observations. The best strategy depends on the intended use, whether it is getting more confidence or accuracy. For a given number of candidate objects (127, equal to 5% of the dataset) for taking spectra, we improve 37% class prediction accuracy as opposed to 20% of a non-naive (non-random) best base-line strategy. Our approach provides a general framework for follow-up strategies and can be extended beyond classification and to include other forms of follow-ups beyond spectroscopy.

show abstract

Section: Autoencodersmentioning

confidence: 99%

An Information Theory Approach on Deciding Spectroscopic Follow-ups

Astudillo

Protopapas

Pichara

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Weighted Approximate-Rank Pairwise Loss using Matrix Factorization (WARP-MF) [35], which maximizes the rank of positive examples by sampling negative examples until a rank violation occurs. And nally, a Variational Autoencoder with multinomial likelihood (Multi-VAE) [16], a deep learning model that extends variational autoencoders. Each base recommendation algorithm (WARP-MF, BPR-MF, and Multi-VAE) was implemented to identify a set of N candidate recommendations.…”

Section: Preference Calculation: User Rating Trainingmentioning

confidence: 99%

One Size Does Not Fit All: Modeling Users' Personal Curiosity in Recommender Systems

Abbas,

Niu

2019

Preprint

View full text Add to dashboard Cite

Today's recommender systems are criticized for recommending items that are too obvious to arouse users' interest. at's why the recommender systems research community has advocated some "beyond accuracy" evaluation metrics such as novelty, diversity, coverage, and serendipity with the hope of promoting information discovery and sustain users' interest over a long period of time. While bringing in new perspectives, most of these evaluation metrics have not considered individual users' di erence: an open-minded user may favor highly novel or diversi ed recommendations whereas a conservative user's appetite for novelty or diversity may not be that large. In this paper, we developed a model to approximate an individual's curiosity distribution over di erent levels of stimuli guided by the well-known Wundt curve in Psychology. We measured an item's surprise level to assess the stimulation level and whether it is in the range of the user's appetite for stimulus. We then proposed a recommendation system framework that considers both user preference and appetite for stimulus where the curiosity is maximally aroused. Our framework di ers from a typical recommender system in that it leverages human's curiosity to promote intrinsic interest with the system. A series of evaluation experiments have been conducted to show that our framework is able to rank higher the items with not only high ratings but also high response likelihood. e recommendation list generated by our algorithm has higher potential of inspiring user curiosity compared to traditional approaches. e personalization factor for assessing the stimulus (surprise) strength further helps the recommender achieve smaller (be er) inter-user similarity.

show abstract

“…In settings where we are not primarily interested in inducing disentangled representations, the β-VAE objective has also been used with β < 1 in order to improve the quality of reconstructions [Alemi et al, 2016, Engel et al, 2017, Liang et al, 2018. While this also decreases the relative weight of to restriction of the mutual information 2 to a subset of "Anchor" variables z a .…”

Section: Related Workmentioning

confidence: 99%

Structured Disentangled Representations

Esmaeili,

Wu,

Jain

et al. 2018

Preprint

View full text Add to dashboard Cite

Deep latent-variable models learn representations of high-dimensional data in an unsupervised manner. A number of recent efforts have focused on learning representations that disentangle statistically independent axes of variation by introducing modifications to the standard objective function. These approaches generally assume a simple diagonal Gaussian prior and as a result are not able to reliably disentangle discrete factors of variation. We propose a two-level hierarchical objective to control relative degree of statistical independence between blocks of variables and individual variables within blocks. We derive this objective as a generalization of the evidence lower bound, which allows us to explicitly represent the trade-offs between mutual information between data and representation, KL divergence between representation and prior, and coverage of the support of the empirical data distribution. Experiments on a variety of datasets demonstrate that our objective can not only disentangle discrete variables, but that doing so also improves disentanglement of other variables and, importantly, generalization even to unseen combinations of factors.

show abstract

Variational Autoencoders for Collaborative Filtering

Cited by 12 publications

References 23 publications

An Information Theory Approach on Deciding Spectroscopic Follow-ups

An Information Theory Approach on Deciding Spectroscopic Follow-ups

One Size Does Not Fit All: Modeling Users' Personal Curiosity in Recommender Systems

Structured Disentangled Representations

Contact Info

Product

Resources

About