IEEE Day Technical Activities, AI Bootcamp #2

#IEEE #ISIMS #SB #New #StudentMember #StudentActivities #isims #TunisiaSection #SMC#SB #Membership #Benefits #ieeeday #AI #Artificial_Intelligence #technical_activities #machine_learning

The second session started with a follow-through of the concepts shared in the previous session. It was led by the same trainer, starting by greeting participants to make sure interaction is ensured.

The session kicked off with a quick review of what had been covered during the previous camp, then did a deep dive on more advanced AI concepts: object detection, instance segmentation, and a comparison between different types of AI. Participants were introduced to generative AI, machine learning, natural language processing, and computer vision, including discussions on supervised and unsupervised learning, and reinforcement learning.

Subsequently, it went to a practical session on data handling, which can be summarized to include topics such as data cleaning, curation, feature selection, splitting and learning algorithms. He then presented a couple of clustering techniques and showed the trainer original and clustered data, then proceeded with dimensionality reduction.

The session involved questions by the participants, and the trainer had also provided practical examples with the help of the Jupyter Notebook integrated into Visual Studio Code. Examples ranged from algorithms on k-nearest neighbors (k-NN) to other functions used within machine learning models.

A practical CSV workshop was then provided to the participants. The trainer performed some hands-on machine learning techniques, focusing on linear regression and writing several related mathematical formulas and programming functions. He explained how to calculate cost functions, regression metrics, and residuals.

The participants were advised to use Anaconda because it is an all-inclusive tool in terms of various libraries and software, or to work with Google Colab on the cloud. They will be working within Python using libraries like "numpy", "Pandas", "Matplotlib", and "Seaborn". The session wrapped up with hands-on exercises left and further Q&A for remaining questions.