Workshop: Deep Learning on Signals

Monday, November 26, 15:00-17:00

Cost: Free

Machine/Deep Learning is a powerful technique for solving complex modeling problems across a broad range of industries. The benefits of deep learning are being realized in applications everywhere, including predictive maintenance, health monitoring, financial portfolio forecasting, and advanced driver assistance etc.

However, Identifying and extracting the relevant features for developing predictive models on sensor data is not a trivial task. Traditional techniques of Identifying the right features and building a model may involve significant domain expertise and is also time consuming. In many applications, lack of having a decent amount of training data also puts a constraint on developing invariant models for signal classification. Moreover, there is an increasing need for developing smart sensor signal processing algorithms which can be either deployed on edge nodes / embedded devices or on the cloud depending on the application.

Join us to learn more about how latest capabilities in MATLAB enable to you to perform Deep learning on signals quickly and with great ease. We will showcase some latest techniques in Deep Learning that let you build models for classification automatically. We will explore a couple of workflows for signal classification using techniques that capture deep insights from signals. You don’t need to have any background in signal processing to use these techniques.

MATLAB based Topics/Examples Include:

  • Deep Learning Techniques applied towards Music Genre Classification - Example
  • Explore Transfer learning workflows to develop predictive models on sensor data using sharp time-frequency representations – Example
  • Explore easy to use signal pre-processing techniques to increase signal quality
  • Explore latest datatypes such as tall arrays to work with data that does not fit in memory alleviating the need for writing special code to work with large sensor data,
  • Leverage high-performance computing resources, such as multicore computers, GPUs, computer clusters to scale up the performance

Bio:

Kirthi K. Devleker is a Sr. Product Manager at MathWorks focusing on Machine Learning and Deep Learning applications for sensor data. Kirthi specializes in helping the scientific community see the benefits of latest advanced signal processing techniques / algorithms to obtain insights from sensor data directly influencing and evolving the signal processing product roadmaps to cover the entire workflow. The capabilities in signal processing products serve the needs of multiple industry verticals such as Medical Devices, Aero-Defense, Automotive and other industries. Kirthi has been with MathWorks for 8 years; and has a master’s in electrical engineering from San Jose State University, CA USA. Prior to joining MathWorks, Kirthi worked as a software evangelist developing sensor characterization tools in MATLAB.