Week 1
Meeting 3
(June 16th, 2022)
GSOC 2022 weekly update
Attendees
- Shaheem Azmal M MD
- Vivek Kumar
- Sahil Jha
- Avinal Kumar
- Sushant Kumar
- Feng Wenhan
- Rohit Pandey
- Thanvi Lahari Pendyala
- Krishna Mahato
- Soham Banerjee
Discussions
- Was majorly working on model training and improving the accuracy. Last week I was geeting the accuracy of 56% for logistic regression model, now after some modification model accuracy is increased to 59%.
- Also tried some other supervised learning models like RandomForest and lsvm.
- And had a discussion with mentors regarding where the training scripts should be kept and we came to conclusion that the model should be trained in our Minerva repo itself and Atarashi will just have to import the model binary from there.
Conclusion and Further Plans
- Will explore some semantic similarity models for text classification
- Also try to improve accuracy further