2023 COSS Data Preparation

This is an introductory workshop to Data Analytics. It starts by introducing the Data Analytics pipeline and its processes. Then, it discusses the different statistical and visualization approaches for conducting Exploratory and Descriptive Analytics on data to answer the question of “What happened in the past?”. The workshop then dives into the art of Data Preparation covering data cleaning, missing values handling, outlier detection and handling, feature transformation and feature engineering.

This workshop will be delivered online in one session:

  • June 19 from 9:00 A.M. to 12:00 P.M. (Eastern Daylight Time)

2023 COSS oneAPI Library and Programming Model for Image Inferencing for Both CPU and GPU

OneAPI is an open standard for a unified application programming interface intended to be used across different compute accelerator architectures, including CPUs, GPUs, AI accelerators and field-programmable gate arrays (FPGA).  It's aim is to unify the programming model as well as simplifying cross-architecture development. It also provides libraries for:

  • Deep Neural Network learning applications.
  • Collective Communications  for machine learning and deep learning projects.
  • Data Analytics making big data analysis faster using optimized algorithms.

In this workshop, we will explore some of the optimized Python math libraries and image inferencing toolkit using oneAPI.  We will also look at how the toolkit can be used to optimize Pytorch and recent YOLOv8 models.  Some performance benchmarking will also be discussed and demonstrated.

This workshop will be delivered online in one session:

  • June 19 from 1:30 P.M. to 4:30 P.M. (Eastern Daylight Time)

2023 COSS Artificial Neural Networks (Deep Learning)

This two-day course will introduce neural network programming concepts, theory and techniques. The class material will begin at an introductory level, intended for those with no experience with neural networks, eventually covering intermediate-to-advanced concepts. The programming language will be Python 3.10; experience with Python programming will be assumed. The Keras neural network framework will be used for neural network programming; no experience with Keras will be expected.

This workshop will be delivered online in four sessions over two days:

  • June 20 from 9:00 A.M. to 12:00 P.M. Eastern Daylight Time
  • June 20 from 1:30 P.M. to 4:30 P.M. Eastern Daylight Time
  • June 21 from 9:00 A.M. to 12:00 P.M. Eastern Daylight Time
  • June 21 from 1:30 P.M. to 4:30 P.M. Eastern Daylight Time

2023 COSS Machine Learning

Machine learning is a subfield of artificial intelligence (AI) that enables computers to learn models from data in order to perform tasks like classifications, recognitions, detections, etc . This is an introductory machine learning course aimed to help the audience build a solid foundation for developing AI applications and exploring more advanced topics. 

The course will begin with an overview of machine learning and its applications. We will then focus on several popular machine learning methods, such as linear regression, decision trees, random forests, and neural networks, and explain how they work and when to use them. Some essential topics like overfitting, regularization, bias-variance trade-off, model evaluation will be addressed in the course. 

As the goal to help the audience to obtain practical skills in machine learning, we will run a list of hands-on exercises throughout the course to illustrate how to apply the aforementioned knowledge to solve real-world problems. The audience will have the opportunity to try some of the exercises on our clusters.

Prerequisites: Beginner’s level of Python is required. Knowledge/experience with Scikit-learn and Tensorflow are preferred but not required.

This workshop will be delivered online in two sessions:

  • June 22 from 9:00 A.M. to 12:00 P.M. Eastern Daylight Time
  • June 22 from 1:30 P.M. to 4:30 P.M. Eastern Daylight Time

2023 COSS Text Mining

Text mining is the process of extracting meaning, patterns, and trends from unstructured textual data. Massive amounts of unstructured text are prevalent in today. Traditional machine learning algorithms handle only numerical or categorical data. Existing data analytical platforms provide special components to facilitate the analysis of textual data. This workshop introduces the topic of text mining and provides a tour with hands-on exercises and demonstrations of four texting mining tools, each of which supports an interesting and diverse set of features.

This workshop will be delivered online in one session:

  • June 23 from 9:00 A.M. to 12:00 P.M