Why Kolabtree

Getting started is quick and easy. No upfront fees

It’s free to request a service and invite bids from experts

Discuss requirements with the expert in detail before accepting statement of work from Kolabtree

Collaborate with the expert directly to get your work done the right way

Fund project when you hire the expert, but approve the deliverables only once work is done

Contact Susovan P.

Contact

Why Kolabtree

Want to hire this expert for a project? Request a quote for free.

Profile Details

Create Project

★★★★★

☆☆☆☆☆

USD 36 /hr

★★★★★

☆☆☆☆☆

Hire Dr. Susovan P.

France

USD 36 /hr

Pure mathematician turned machine learning researcher looking for math and machine learning related projects

Profile Summary

Machine learning and statistics skills:

Classification and regression algorithms (SVM, LDA, LR etc.), cross validation, regularization, kernel methods, ensemble methods: bagging and boosting, dimensionality reduction - PCA, variance threshold.

Clustering: K-means, DBSCAN.

Deep learning: Multilayer perceptron (MLP), convolutional and recurrent neural networks(CNN, RNN), long short term memoery (LSTM).

Sequential models: Hidden Markov Models (HMM), RNN.

Advanced probability, statstical tests-hypothesis testing, inference.

Advanced mathematics skills:

Differential geometry, algebraic topology, complex analysis and Riemann surfaces, real analysis, field and Galois theory, partial differential equations, measure theory, abstract and linear algebra, functional analysis.

Programming ans software skills:

Python, Matlab, objected oriented programming, NumPy, sk-learn, Pandas, Seaborn, OpenCV, audio signal processing libraries (pydub, pyAudioAnalysis), elementary familiarity with TensorFlow, Linux, LaTeX, Git.

Projects:

1) Audio signal processing:  Music and speech detection (coding in Python and TensorFlow): Developed a kernelized SVM classifier that can classify music and speech with 98% accuracy using the MFCC features of the audios, and used it to segment recorded call audios into music and speech. Also developed a long short term memory (LSTM) classifier that gave 95% accuracy on similar data.

2) Geolocation data analysis: Determining popular places from GPS and other geolocation related data (coding in Python): constructed a suitable cost function which is an indicator of how popular a co-ordinate of earth's surface is, in the sense that it measures how many trajectories pass at or near that co-ordinate or through the corresponding geohash. By minimizing this cost function, which is further a function of distance to trajectories, I was able to estimate the popularity of a place.

3) Academic research on classification of medical data: Classification of fetal alcohol syndrome patients (coding in MATLAB): Performed linear discriminant analysis (LDA) on 175 precentral and central sulcal curves, with a 120-dimensional feature space consisting of Fourier coefficients of the curves. Estimated the performance of this classifier by a 10-fold cross validation to obtain classification accuracy of 76%.

Subject Matter Expertise

Services

Data & AI

Data & AI Statistical Analysis, Algorithm Design-Non ML

Work Experience

R&D data scientist in Audio Signal Processing

Batvoice

April 2019 - July 2019

Data Scientist

Connecthings

July 2018 - July 2018

Education

PhD in Mathematics

Rutgers University

January 2019 - January 2019

Certifications

Certification details not provided.