IBM - Data Scientist - Cognitive Computing

Qualification : Computer Science, Engineering, Statistics, Mathematics, MBA, Economics, Post Graduate Degree preferred.

Mandatory :

Quantitative :

- Skills to analyze data, and interpret analysis results in business in terms. This includes logical thinking, reasoning, hypothesis formulation, evidence-based evaluation, quantitative approximation and intuitive cross-validation.

- Ability to create scientific models for solving business problems from first principles is mandatory. Solid background of school-level mathematics is an absolute imperative.

- Formal training in, or experiential learning of, statistical methods and techniques would be necessary.

Machine Learning :

- Skills to develop programmatic interventions to generate insights and develop cognitive systems.

- This includes skills to translate business problems to machine learning needs, design and develop machine learning systems for prediction, prescription. optimization and fact-based decisions, and tweak and train the models for accuracy improvement and automated self-adjustment.

It is required to have excellent programming skill in any one of the below languages, or good skills in any two languages from the following groups :

Object-Oriented : Java / C++

Procedural : Python / Perl / C/ Matlab / Octave / Fortran / Pascal

Special Purpose : R / SAS / SPSS / Matlab / Octave / Rapid Miner / Weka / Mahout / Spark / LISP

In addition, good understanding of feedback learning (e.g. reinforcement learning, Q-Learning, heuristic algorithms) would be helpful.

- Information Management : Skills to extract, transform, prepare and manipulate large datasets and unstructured content using standard RDBMS / SQL or big data and computing platforms like Hadoop / NoSql / Hbase / Pig / Hive.

Preferred Skills :

- Natural Language Processing

- Image / Video Processing

- Streaming Data Analytics

- Artificial Intelligence and Cognitive Systems

- Human Computer Interaction / Visualization