Nikos D. Sidiropoulos

Translation Invariant Word Embeddings

This work focuses on the task of finding latent vector representations of the words in a corpus. In particular, we address the issue of what to do when there are multiple languages in the corpus. Prior work has, among other techniques, used canonical correlation analysis to project pre-trained vectors in two languages into a common space. We propose a simple and scalable method that is inspired by the notion that the learned vector representations should be invariant to translation between languages. We show empirically that our method outperforms prior work on multilingual tasks, matches the performance of prior work on monolingual tasks, and scales linearly with the size of the input data (and thus the number of languages being embedded).

A General Optimization Framework for Stochastic Routing in Wireless Multi-hop Networks

We introduce a novel approach to routing based on the so called pairwise packet delivery ratio matrix whose entries represent the probability that a given user decodes the packet transmitted by any other user. We show that this leads naturally to a model in which routing algorithms are described by the evolution of a Markov chain enabling the definition of deliverability criteria in terms of absorbing states. We further introduce optimal routing protocols by selecting the routing matrix from a convex polygon containing all feasible routing matrices. The criteria of optimality include minimization of the packet error probability for a given delay bound and the minimization of the average packet delay. These metrics are correspondingly meaningful in the context of real time transmissions - e.g., voice and/or video - and delay insensitive data - e.g., file transfers.

Increasing the throughput of spread-Aloha protocols via long PN spreading codes

Random access Aloha protocols have well documented merits in terms of simplicity and favorable delay-throughput trade-off under moderate bursty traffic loads. Short spreading codes have been used in conjunction with random access to endow Aloha with benefits originating from spread-spectrum communications. Instead of short, symbol-periodic spreading, this paper considers long pseudo-random (PN) packet-periodic sequences in the context of spread-Aloha and establishes that long PN codes increase the maximum stable throughput by reducing the probability of collisions. Relying on a dominant system approach, we analyze the resultant throughput and demonstrate that increasing the PN code length quickly transforms the collision-limited channel to an interference-limited one. In particular, we investigate how throughput depends on user load and packet length. Finally, we discuss synchronization issues and provide corroborating numerical results.

Sparse dictionary learning from 1-BIT data

This work examines a sparse dictionary learning task - that of fitting a collection of data points, arranged as columns of a matrix, to a union of low-dimensional linear subspaces - in settings where only highly quantized (single bit) observations of the data matrix entries are available. We analyze a complexity penalized maximum likelihood estimation strategy, and obtain finite-sample bounds for the average per-element squared approximation error of the estimate produced by our approach. Our results are reminiscent of traditional parametric estimation tasks - we show here that despite the highly-quantized observations, the normalized per-element estimation error is bounded by the ratio between the number of “degrees of freedom” of the matrix and its dimension.

NMF revisited: New uniqueness results and algorithms

Non-negative matrix factorization (NMF) has found numerous applications, due to its ability to provide interpretable decompositions. Perhaps surprisingly, existing results regarding its uniqueness properties are rather limited, and there is much room for improvement in terms of algorithms as well. Uniqueness and computational aspects of NMF are revisited here from a geometrical point of view. Both symmetric and asymmetric NMF are considered, the former being tantamount to element-wise non-negative square-root factorization of positive semidefinite matrices. New and insightful uniqueness results are derived, e.g., it is shown that a sufficient condition for uniqueness is that the conic hull of the latent factors is a superset of a particular second-order cone. Checking this is shown to be NP-complete; yet it offers insights on latent sparsity, as is also shown in a new necessary condition, to a smaller extent. On the computational side, a new efficient algorithm for symmetric NMF is proposed which uses Procrustes rotations. Simulation results show promising performance with respect to the state-of-art. The new algorithm is also applied to a clustering problem for co-authorship data, yielding meaningful and interpretable results.

Time-varying fair queuing scheduling for multicode CDMA based on dynamic programming

Fair queuing (FQ) algorithms, which have been proposed for QoS wireline-wireless networking, rely on the fundamental idea that the service rate allocated to user m is proportional to a positive weight /spl phi//sub m/. Targeting wireless data networks with a multicode CDMA-based physical layer, we develop FQ with time-varying weight assignments in order to minimize the queuing delays of mobile users. Applying dynamic programming, we design a computationally efficient algorithm which produces the optimal service rates while obeying (i) constraints imposed by the underlying physical layer, and (ii) QoS requirements. Simulations illustrate the merits of our designs.

Efficient and effective nearest neighbor search in a medical image database of tumor shapes

We examine the problem of finding similar tumor shapes. The main contribution of this work is the proposal of a natural similarity function for shape matching called the ‘morphological distance ’. This function has two desirable properties: (a) it matches human perception of similarity, as we illustrate with precision/recall experiments; (b) it can be lower-bounded by a set of features, leading to fast indexing for range queries and nearest neighbor queries.