Portfolio - partial non-proprietary repository

Only non-proprietary projects and helper code used to implement these projects including:

NLP

(Python)

• Word embeddings including contextual embeddings

• Natural language understanding and word sense disambiguation

• Sentiment classifier for customer reviews

• Custom part-of-speech tagger

• Text classification solutions using Word2vec, sci-Kit learn, Metapy and a number of other tools

• Document similarity solution for a large collection of documents to search for duplicates, near-duplicates, and versions of documents

• Batchers sending multiple REST API calls to a text mining engine (to test a document classification model, extract key words from text, parse log files)

• Comprehensive customer review analysis consisting of 7 distinct steps from topic modeling to a web app

• Information extraction from a large AWS public dataset of material safety data sheets for subsequent use in industrial safety recommendations

• Apache storm topologies for NLP

• Topic modeling

Machine Learning

(Python and R):

• Image Classification and Analysis Including Denoising and Segmentation Using A Number of Different Tools (several projects)

• Linear and Logistic Regressions

Cloud Technologies

(Python and Java):

• Hadoop MapReduce

• Apache Spark MLlib and Giraph

Data Mining

(Python)

• Implementation of Several Data Mining Algorithms

Visualization

• Website at https://agnedil.github.io/; code at https://github.com/agnedil/agnedil.github.io - a D3.js Visualization Project

• Various Visualization Methods Used in All Projects in This Repository Including D3.js, Matplotlib, Ggplot2, Tableau, Spotfire