Since the last Nuit Blanche in Review (April 2014), Nuit Blanche crossed the 1 million visits (and it is about to get close to 3 million page views). Yesterday's entry on the slides of the Sublinear Algorithms 2014 meeting still resonates with me, but this past month, I also created a meetup around note by note cooking ( Connections between Note by Note Cooking and Machine Learning ). The Advanced Matrix Factorization Jungle page added some factorizations found in Graph Matching and Archetypal Analysis. We also wondered how FROG could be addressed with Nonlinear Compressive Sensing, saw the growth of Multiple Regularizers, but also saw the possibility of one-bit compressive sensing to provide some way for understanding neural networks or how real Neurons could be thought as a Signal Processing Device. Sharp phase transition such as Donoho-Tanner was also seen as a way to probe good from bad neural networks in Compressive Sensing and Short Term Memory / Visual Nonclassical Receptive Field Effects or provide Sharp Performance Bounds for Graph Clustering . We were also made aware of the connection how the theory of convex optimization influences Machine Learning or the geometric perspective of the problem of the estimation in high dimensions, we also had an overview of Sparsity-Aware Learning and Compressed Sensing.

We also had quite a few implementations made available.

Implementations:

- Fast and Robust Archetypal Analysis for Representation Learning
- Probabilistic Archetypal Analysis
- On the Convergence of Approximate Message Passing with Arbitrary Matrices
- CGIHT, ASD and ScaledASD: Compressed Sensing and Matrix Completion
- GAGA: GPU Accelerate Greedy Algorithms for Compressed Sensing - new version -
- PR-GAMP : Compressive Phase Retrieval via Generalized Approximate Message Passing
- PSPG: Efficient Algorithms for Robust and Stable Principal Component Pursuit Problems
- Image Restoration Using Joint Statistical Modeling in Space-Transform Domain
- Group-based Sparse Representation for Image Restoration
- SPAL: Sparse Projection-based Adaptive Learning
- ADMIP: An Alternating Direction Method with Increasing Penalty for Stable Principal Component Pursuit
- One-bit compressive sensing with norm estimation
- PNOPT: Proximal Newton-type methods for minimizing composite functions
- OptShrink: An algorithm for improved low-rank signal matrix denoising by optimal, data-driven singular value shrinkage
- SUGAR : Stein Unbiased GrAdient estimator of the Risk (SUGAR) for multiple parameter selection
- emd_flow : The Constrained Earth Mover Distance Model,with Applications to Compressive Sensing
- Tree_approx : A Fast Approximation Algorithm for Tree-Sparse Recovery
- FCT : Fast Compressive Tracking
- WNNM: Weighted Nuclear Norm Minimization with Application to Image Denoising
- GHP : Gradient Histogram Estimation and Preservation for Texture Enhanced Image Denoising
- Selecting thresholding and shrinking parameters with generalized SURE for low rank matrix estimation
- Image Compressive Sensing Recovery Using Adaptively Learned Sparsifying Basis via L0 Minimization
- WESNR : Mixed Noise Removal by Weighted Encoding with Sparse Nonlocal Regularization
- Fast Tracking via Spatio-Temporal Context Learning
- RCoS : Image Compressive Sensing Recovery via Collaborative Sparsity
- Random Projections for Non-negative Matrix Factorization

Sunday Morning Insights:

Experiment /Hardware

Focused Entries:

- Sparsity-Aware Learning and Compressed Sensing: An Overview
- Graph Matching: Relax at Your Own Risk
- Addressing FROG with Nonlinear Compressive Sensing
- Multiple Regularizers: Multi-View Learning and Hyperspectral Imagery
- One-bit compressive sensing with norm estimation
- Theory of Convex Optimization for Machine Learning / Estimation in high dimensions: a geometric perspective
- A Neuron as a Signal Processing Device
- Compressive Sensing and Short Term Memory / Visual Nonclassical Receptive Field Effects
- Sharp Performance Bounds for Graph Clustering via Convex Optimization

Meetings / Meetups :

- CfP: 2nd Large-Scale Recommender Systems and RecSysTV
- Ce Soir / Tonight: Paris Machine Learning Meetup #11: Learning What Is It Good For ? SPARFA, Learning to Interact, Action recognition with CNNs and Prediction APIs
- Paris Machine Learning Meetup #11: Learning, What Is It Good For ?
- Registration: 2014 Workshop on Algorithms for Modern Massive Data Sets (MMDS 2014)

Q&A:

Video

Comics

Jobs:

Hamming's time

Saturday Morning Videos

**Join the CompressiveSensing subreddit or the Google+ Community and post there !**

Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.