skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: An Estimation and Analysis Framework for the Rasch Model
The Rasch model is widely used for item response analysis in applications ranging from recommender systems to psychology, education, and finance. While a number of estimators have been proposed for the Rasch model over the last decades, the associated analytical performance guarantees are mostly asymptotic. This paper provides a framework that relies on a novel linear minimum mean-squared error (L-MMSE) estimator which enables an exact, nonasymptotic, and closed-form analysis of the parameter estimation error under the Rasch model. The proposed framework provides guidelines on the number of items and responses required to attain low estimation errors in tests or surveys. We furthermore demonstrate its efficacy on a number of real-world collaborative filtering datasets, which reveals that the proposed L-MMSE estimator performs on par with state-of-the-art nonlinear estimators in terms of predictive performance.  more » « less
Award ID(s):
1652065
PAR ID:
10082581
Author(s) / Creator(s):
; ;
Date Published:
Journal Name:
Proceedings of the 35th International Conference on Machine Learning
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; (, 2018 Annual American Control Conference (ACC))
    Monitoring of a linear diffusive network dynamics that is subject to a stationary stochastic input is considered, from a graph-theoretic perspective. Specifically, the performance of minimum mean square error (MMSE) estimators of the stochastic input and network state, based on remote noisy measurements, is studied. Using a graph-theoretic characterization of frequency responses in the diffusive network model, we show that the performance of an off-line (noncausal) estimator exhibits an exact topological pattern, which is related to vertex cuts and paths in the network's graph. For on-line (causal) estimation, graph-theoretic results are obtained for the case where the measurement noise is small. 
    more » « less
  2. ; (, 2017 51st Asilomar Conference on Signals, Systems, and Computers)
    Partially-observed Boolean dynamical systems (POBDS) are large and complex dynamical systems capable of being monitored through various sensors. However, time, storage, and economical constraints may impede the use of all sensors for estimation purposes. Thus, developing a procedure for selecting a subset of sensors is essential. The optimal minimum mean-square error (MMSE) POBDS state estimator is the Boolean Kalman Filter (BKF) and Smoother (BKS). Naturally, the performance of these estimators strongly depends on the choice of sensors. Given a finite subsets of sensors, for a POBDS with a finite observation space, we introduce the optimal procedure to select the best subset which leads to the smallest expected mean-square error (MSE) of the BKF over a finite horizon. The performance of the proposed sensor selection methodology is demonstrated by numerical experiments with a p53-MDM2 negative-feedback loop gene regulatory network observed through Bernoulli noise. 
    more » « less
  3. ; ; ; (, International Symposium on Information Theory)
    null (Ed.)
    The minimum mean-square error (MMSE) achievable by optimal estimation of a random variable S given another random variable T is of much interest in a variety of statistical contexts. Motivated by a growing interest in auditing machine learning models for unintended information leakage, we propose a neural network-based estimator of this MMSE. We derive a lower bound for the MMSE based on the proposed estimator and the Barron constant associated with the conditional expectation of S given T . Since the latter is typically unknown in practice, we derive a general bound for the Barron constant that produces order optimal estimates for canonical distribution models. 
    more » « less
  4. ; ; ; ; (, IEEE International Symposium on Information Theory (ISIT))
    This paper deals with linear equalization in massive multi-user multiple-input multiple-output (MU-MIMO) wireless systems. We first provide simple conditions on the antenna configuration for which the well-known linear minimum mean-square error (L-MMSE) equalizer provides near-optimal spectral efficiency, and we analyze its performance in the presence of parameter mismatches in the signal and/or noise powers. We then propose a novel, optimally-tuned NOnParametric Equalizer (NOPE) for massive MU-MIMO systems, which avoids knowledge of the transmit signal and noise powers altogether. We show that NOPE achieves the same performance as that of the L-MMSE equalizer in the large-antenna limit, and we demonstrate its efficacy in realistic, finite-dimensional systems. From a practical perspective, NOPE is computationally efficient and avoids dedicated training that is typically required for parameter estimation. 
    more » « less
  5. ; ; ; (, arXiv)
    Addressing the fundamental challenge of signal estimation from noisy data is a crucial aspect of signal processing and data analysis. Existing literature offers various estimators based on distinct observation models and criteria for estimation. This paper introduces an innovative framework that leverages topological and geometric features of the data for signal estimation. The proposed approach introduces a topological tool -- {\it peak-persistence diagram} (PPD) -- to analyze prominent peaks within potential solutions. Initially, the PPD estimates the unknown shape, incorporating details such as the number of internal peaks and valleys. Subsequently, a shape-constrained optimization strategy is employed to estimate the signal. This approach strikes a balance between two prior approaches: signal averaging without alignment and signal averaging with complete elastic alignment. Importantly, the proposed method provides an estimator within a statistical model where the signal is affected by both additive and warping noise. A computationally efficient procedure for implementing this solution is presented, and its effectiveness is demonstrated through simulations and real-world examples, including applications to COVID rate curves and household electricity consumption curves. The results showcase superior performance of the proposed approach compared to several current state-of-the-art techniques. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email