As a teaching assistant I was in charge of the following activities:

• Review exams taken by students.
• Design and create tests and projects for students.
• Give assistantships to students to clarify their doubts.
• Analyze and pre-process time series of CPU load.
• Create forecasting models with LSTM and CNN.
• Analyze and preprocess magnetics resonance imagges (MRI).
• Classification and segmetation of regions on MRIs.

As a research assistant I was in charge of the following activities:

• Research and definition of pipelines for data preprocessing for timelines and 3D images and configuration of NVIDIA GPU training environments and use of high-end computers (CEDIA) through jupyter notebooks.
• Data visualization of pre-processed data through libraries such as matplotlib and seaborn. Real-time recording and tracking of training performance through personally-authored callbacks, to be accessible on Telegram and Webhooks. Saving of metrics and model performance with Tensorboard for later analysis.
• Creation of LSTM based model to predict the behavior of CPU usage in Telconet's datacenter in Ecuador, from time series of the percentage of CPU usage recorded to trigger early warnings according to a specified limit.
• Creation of Transformers based model to segment the regions of the substantia nigra and subthalamic nucleus from T1-weighted MRI. In order to provide a model to assist in possible interventions to help in the treatment of Parkinson's disease.
• Creation of Transformers based model to find and classifiy T1-weighted MRI to patterns of mental disorder associated with Parkinson's disease.

Tech used

• Tensorflow
• Jupyter notebook
• Matplotlib
• Seaborn
• CUDA
• Java
• Python