ADAjay Deshpande

Ajay Deshpande

*This is an algorithmically generated author record based on publicly available information.

Ajay Deshpande has a total of 15 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    On Machine Learning-Based Short-Term Adjustment of Epidemiological Projections of COVID-19 in US

    by Sarah Kefayati, Fred Roberts, Sayali Pethe, Xuan Liu, Hu Huang, Vishrawas Gopalakrishnan, Piyush Madan, Jianying Hu, Prithwish Chakraborty, Raman Srinivasan, et al.

    Epidemiological models have provided valuable information for the outlook of COVID-19 pandemic and relative impact of different mitigation scenarios. However, more accurate forecasts are often needed at near term for planning and staffing. We present our early results from a systemic analysis of short-term adjustment of epidemiological modeling of COVID 19 pandemic in US during March-April 2020. Our analysis includes the importance of various types of features for short term adjustment of the predictions. In addition, we explore the potential of data augmentation to address the data limitation for an emerging pandemic. Following published literature, we employ data augmentation via clustering of regions and evaluate a number of clustering strategies to identify early patterns from the data. From our early analysis, we used CovidActNow as our underlying epidemiological model and found that the most impactful features for the one-day prediction horizon are population density, workers in commuting flow, number of deaths in the day prior to prediction date, and the autoregressive features of new COVID-19 cases from three previous dates of the prediction. Interestingly, we also found that counties clustered with New York County resulted in best preforming model with maximum of R 2 = 0.90 and minimum of R 2 =0.85 for state-based and COVID-based clustering strategy, respectively.


  • Open Access

    Machine learning based co-creative design framework

    by Brian Quanz, Wei Sun, Ajay Deshpande, Dhruv Shah, Jae-Eun Park

    We propose a flexible, co-creative framework bringing together multiple machine learning techniques to assist human users to efficiently produce effective creative designs. We demonstrate its potential with a perfume bottle design case study, including human evaluation and quantitative and qualitative analyses.

  • Open Access

    Guaranteed sensor coverage with the weighted-$D^2$ sampling

    by Ajay Deshpande

    In this paper we focus on the mobile sensor coverage problem formulated as a continuous locational optimization problem. Cort\`es et al. first proposed a distributed version of the Lloyd descent algorithm with guaranteed convergence to a local optima. Since then researchers have studied a number of variations of the coverage problem. The quality of the final solution with the Lloyd descent depends on the initial sensor configuration. Inspired by the recent results on a related $k$-means problem, in this paper we propose the weighted-$D^2$ sampling to choose the initial sensor configuration and show that it yields $O(\log k)$-competitive sensor coverage before even applying the Lloyd descent. Through extensive numerical simulations, we show that the initial coverage with the weighted-$D^2$ sampling is significantly lower than that with the uniform random initial sensor configuration. We also show that the average distance traveled by the sensors to reach the final configuration through the Lloyd descent is also significantly lower than that with the uniform random configuration. This also implies considerable savings in the energy spent by the sensors during motion and faster convergence.