Analysed Neurophysical data which was recorded using an EEG for more than 900 children. Employed different ML models such as Random Forest, SVM etc. with multicollinearity on EEG features while trying to predict the BSID scores of those children. 1. Followed a three-step approach: Data cleaning, Finding features which best correlate with BSID(Bayley Scales of Infant Development) and predicting BSID scores using derived variables or features.2. The neurophysical data consisted of EEG features such as alpha, delta, gamma, theta wave intensity and a few derived variables such as complexity, Amplitude Standard Deviation, etc., which were incorporated after a suitable literature review.3. Pearson correlation was then used to assess the best-performing variables.4. Different combinations of features/variables were used finally with multicollinearity to predictthe BSID scores.