Bikash Joshi, PhD Student

I am a PhD student at AMA team in the LIG (Laboratoire d'Informatique de Grenoble) Lab. My PhD thesis is supervised by Prof. Massih-Reza Amini . and Dr. Franck Iutzeler. I am a data enthusiast and my career aspiration is to become a leading researcher in the field of Machine Learning and Data Mining.

PhD Thesis

My PhD thesis is about developing scalable algorithms for large-scale machine learning problems. I have worked on two projects. Firstly, I worked on reducing Multiclass classification problems to binary classification problems. The proposed reduction technique has significant performance benefits especially for large-scale multiclass classification. Secondly, I worked on distributed stochastic optimization with applications to collaborative filtering and binary classification. We proposed an asynchronous framework for this and used advanced distrbuted computing tools (Spark and MPI) for the implementation.

Research Interests

  1. Recommender Systems
  2. Distributed Machine Learning
  3. Large-scale multi-class classification
  4. Distributed Computing and Optimization

Publications

  1. Aggressive Sampling for Multi-class to Binary Reduction with Applications to Text Classification B. Joshi, M.R. Amini, I. Partalas, F. Iutzeler, Y. Maximov. arXiv (2017).
  2. Asynchronous Distributed Matrix Factorization with Similar User and Item Based Regularization. B. Joshi, F. Iutzeler, M.R. Amini. Proceedings of the 10th ACM Conference on Recommender Systems (RecSys'16), Boston, USA.
  3. Software Clone Detection Using Clustering Approach. B. Joshi, P. Budhathoki, W.L. Woon, D. Svetinovic. International Conference on Neural Information Processing (ICONIP'15), Istanbul, Turkey.
  4. On binary reduction of large-scale multiclass classification problems. B. Joshi, M.R. Amini, I. Partalas, L. Ralaivola, N. Usunier, E. Gaussier. International Symposium on Intelligent Data Analysis (IDA'15), Saint-Etienne, France.
  5. Automatic Rooftop Detection Using a Two-stage Classification. B. Joshi, H. Baluyan, A. Al. Hinai, W.L. Woon. UKSIM-AMSS 16th International Conference on Modelling and Simulation (UKSIM), IEEE (2014).
  6. Novel Approach for Rooftop Detection using Support Vector Machine. H. Baluyan, B. Joshi, A.Al. Hinai, W.L.Woon, ISRN Machine Vision Journal (2013).

Talks and Workshops

  1. Talk on "Aggressive Double Sampling for Multi-class to Binary Reduction " at Macaron Workshop (2017), Inria, France.
  2. Talk on "Asynchronous Distributed Matrix Factorization with Similar User and Item Based Regularization" at RecSys Conference (2016), Boston, USA.
  3. Talk on "Multi-class to Binary reduction of Large-scale classification Problems.". Big Multi-Target Prediction @ ECML/PKDD'15, Porto, Portugal (2015) .
  4. Talk on "On Binary Reduction of Large-scale Multiclass Classification Problems" at Spring school on Large-scale learning, Grenoble, France (2014).
  5. Talk on "Rooftop Detection for Planning of Solar PV Deployment: A Case Study in Abu Dhabi" at DARE Workshop@ ECML/PKDD'14, Nancy, France.

Contact