Careers
Algorithm Research Engineer
Description
The work being done at Numenta lies at the intersection of machine learning, neuroscience and software engineering. As a member of the algorithm team, you will put your mathematical and algorithm skills to work on real-life problems involving recognition and prediction of patterns. You will research and develop new probabilistic learning and inference algorithms for Numenta's Hierarchical Temporal Memory (HTM) technology and use those algorithms to solve problems in computer vision, speech recognition, multi-variable time-series prediction etc. In the process, you will build a better understanding of cortical functions and look at machine learning and computational learning theory from novel view points. You will document your inventions for publications and patents, and keep a keen eye on advances in the field of machine learning and neuroscience.
Qualifications and Skills
The primary requirement for this job is that you be a critical thinker and problem solver. You should be well versed in the fundamentals of probability theory, linear algebra, signal processing, optimization and machine learning, and at the same time be able to think beyond the established mathematical frameworks for learning and pattern recognition. You should be able to quickly prototype algorithms and test them in a large problem setting. In addition to being analytical and rigorous about particular problems, you should also be able to synthesize ideas from disparate disciplines. You should be able to enjoy intense discussions and should have a healthy dose of skepticism and rigor.
The ideal candidate will have the following qualifications and skill set. We also will consider strong candidates from related disciplines (eg; communication theory, information theory & coding) who are passionate about learning new techniques and applying their experience to real world problems.
- PhD in Computer Science, Electrical Engineering or a related discipline OR Masters in CS/EE with relevant research work experience
- Strong understanding of Probability Theory
- Course work or experience in Bayesian Networks, Markov Models and Convex Optimization
- Exposure to Bayesian Belief Propagation
- Experience applying learning algorithms to pattern recognition and prediction problems
- Experience working with large data sets
- Knowledge of recent developments in machine learning
- Course work or experience in image/video processing and computer vision
- Exposure to Information Theory and Statistics
- Should be an excellent problem solver and a motivated self-starter
- Excellent communication and documentation skills
- Strong programming skills.
- C/C++ and MATLAB required - Knowledge of Python desirable
Keywords: machine learning, signal processing, information theory, convex optimization, probability theory, algorithm development.
