The IDC CS Colloquium

Meitar Ronen - DeepDPM: Deep Clustering With an Unknown Number of Clusters

Deep Learning (DL) has shown great promise in the unsupervised task of clustering. That said, while in classical (i.e., non-deep) clustering the benefits of the nonparametric approach are well known, most deep-clustering methods are parametric: namely, they require a predefined and fixed number of clusters, denoted by K. When K is unknown, however, using model-selection criteria to choose its optimal value might become computationally expensive, especially in DL as the training process would have to be repeated numerous times.

26/05/2022 - 13:30

Ben Galili - From Stable Statistics to Robust Differentially Expressed Genes

Statistical hypothesis test is a fundamental tool in statistics, where one can determine whether to accept or reject the null hypothesis (the default statement of no difference). Rejecting the null hypothesis means that the probability of getting the observed results is very low - lower than a predetermined significance level. This probability is called a p-value and the lower the p-value is the more significant the results.

19/05/2022 - 13:30

Nati Linial - Graphs as geometric objects

Many of the basic algorithmic questions deal with the geometry of graphs. We know many variations on the theme of computing the metric of edge-weighted graphs. E.g., “all-pairs shortest paths”. As I will try to convey in this lecture, there is much that we know beyond that as well as broad barely charted territories. For example, let G=(V,E) be a graph and let w(e)>0 be given positive real numbers associated with every edge e in E.

12/05/2022 - 13:30

Ran Cohen - Reaching Agreement Without Saying Much: Byzantine Agreement With Polylog Bits Per Party

Byzantine agreement (BA), the task of n parties to agree on one of their input bits in the face of malicious agents, is a powerful primitive that lies at the core of a vast range of distributed protocols. Interestingly, in protocols with the best overall communication, the demands of the parties are highly unbalanced: the amortized cost is polylog(n) bits per party, but some parties must send \Omega(n) bits.

28/04/2022 - 13:30

Shimon Shocken - From Xor to Doom: Pegasus Exposed

Pegasus is a zero-click attack: The target is sent a text message, and that's it – the device is compromised – the message doesn't even have to be read. According to a recent Google report: "Short of not using the device, there is no way to prevent exploitation... Pegasus is a weapon against which there is no defense... This is one of the most technically sophisticated exploits we've ever seen".

24/03/2022 - 13:30

Aviv Yehezkel: Network Anomaly Detection Using Transfer Learning Based on Auto-Encoders Loss Normalization

Anomaly detection in computer networks is a classic, long-term research problem. Previous attempts to solve it have used auto-encoders to learn a representation of the normal behaviour of networks and detect anomalies according to reconstruction loss. In this talk, I'll present the concept of "auto-encoder losses transfer learning".

10/03/2022 - 13:30

Guy Gaziv - Mind Reading: Decoding Visual Experience from Brain Activity

Reconstructing and semantically classifying observed natural images from novel (unknown) fMRI brain recordings is a milestone for developing brain-machine interfaces and for the study of consciousness. Unfortunately, acquiring sufficient “paired” training examples (images with their corresponding fMRI recordings) to span the huge space of natural images and their semantic classes is prohibitive, even in the largest image-fMRI datasets.

11/11/2021 - 13:30