Search Results for author: Milos Manic

Found 7 papers, 1 papers with code

Spintronic Physical Reservoir for Autonomous Prediction and Long-Term Household Energy Load Forecasting

no code implementations6 Apr 2023 Walid Al Misba, Harindra S. Mavikumbure, Md Mahadi Rajib, Daniel L. Marino, Victor Cobilean, Milos Manic, Jayasimha Atulasimha

By comparing our spintronic physical RC approach with state-of-the-art energy load forecasting algorithms, such as LSTMs and RNNs, we conclude that the proposed framework presents good performance in achieving high predictions accuracy, while also requiring low memory and energy both of which are at a premium in hardware resource and power constrained edge applications.

Load Forecasting regression

Self-Supervised and Interpretable Anomaly Detection using Network Transformers

no code implementations25 Feb 2022 Daniel L. Marino, Chathurika S. Wickramasinghe, Craig Rieger, Milos Manic

Monitoring traffic in computer networks is one of the core approaches for defending critical infrastructure against cyber attacks.

Anomaly Detection

Physics Enhanced Data-Driven Models with Variational Gaussian Processes

no code implementations5 Jun 2019 Daniel L. Marino, Milos Manic

Centuries of development in natural sciences and mathematical modeling provide valuable domain expert knowledge that has yet to be explored for the development of machine learning models.

Gaussian Processes Inductive Bias

Building Energy Load Forecasting using Deep Neural Networks

no code implementations29 Oct 2016 Daniel L. Marino, Kasun Amarasinghe, Milos Manic

Experimental results showed that the standard LSTM failed at one-minute resolution data while performing well in one-hour resolution data.

Decision Making energy management +2

Fast Trajectory Simplification Algorithm for Natural User Interfaces in Robot Programming by Demonstration

1 code implementation25 Aug 2016 Daniel L. Marino, Milos Manic

While most current trajectory simplification algorithms are tailored for GPS trajectories, our approach focuses on smooth trajectories for robot programming by demonstration recorded using motion capture systems. Two variations of the algorithm are presented: 1. aims to preserve shape and temporal information; 2. preserves only shape information.

Cannot find the paper you are looking for? You can Submit a new open access paper.