Deep Learning Workshop

A Non-Mathy Deep Learning Workshop – February 5, 2022

For students and professionals to understand the basics and impact of Deep Learning in today’s world


3rd Annual MIT MIMO Deep Learning Workshop
February 3 – 4, 2021

  • 100 Students and Faculty
  • 1 Day Hands-on Python coding
  • 1 Day Faculty and Industry Presentations
  • Focus on time-series data applications of ML
  • NVIDIA Certification

MIT Machine Intelligence for Manufacturing and Operations in conjunction with NVIDIA and AWS are sponsoring the third annual MIT MIMO Deep Learning Workshop Feb 3-4. The Purpose of the 2-Day workshop is to help prepare students to implement real world machine learning / deep learning projects during their thesis work, IAP projects, internships and beyond.

NVIDIA will host a 6 – 7 hour hands-on (Python / Jupyter notebooks) workshop where students will predict failures in time-series data using: XGBoost, LSTM-based NN, 1-D LSTM-based autoencoders. Students will compare model accuracy from the three approaches and address class imbalance issues and efficient hyper parameter turning. Since this is a live code workshop, advanced students can work at their own pace and feel free to experiment with different approaches and hyper-parameter settings as well as complete advanced self-paced exercises.

In addition to the hands-on portion of the lab run by NVIDIA, we will also have case studies and research presented by MIT faculty and alums. The intent of the case studies is to broaden students view on how the many use cases for machine learning in manufacturing and operations.

Videos on Demand:

Deep Learning Workshop Series

  • 2018  – Defect Detection in Biopharmaceuticals
    • Model Thesis:  Amgen – Nyovanie ’19  
    • Sponsor: AWS
    • Key Learnings : Cloud-based training with Edge-based inference

  • 2019 – False Positive Rejection in PCB Assembly
    • Model Thesis: FLEX – Foster ’19 
    • Sponsors: NVIDIA, AWS
    • Key Learnings: Data Curation; Model performance comparison; Data pipelining

  • 2020 – Expand to two-day virtual session
    • False Positive Rejection in PCB Assembly (above) – 1 day
    • LGO Alumni Machine Intelligence Use Case review – 1 Day
    • Key learnings: Broader view of types of problems solved with MI, range of approaches and their applicability, implementation best practices