I did not see updated slides at time of review. I think the research question needs a bit of tweaking for clarity as it is now.
Code:
Your work process was easy to follow. I think your notebook should be condensed for easy reading by anyone wanting to review your portfolio. Extremely long print statements / outputs in cells should be snipped. Glad you were able to leverage beautifulsoup. Make sure you dedicate sufficient space to highlight your custom sentiment function.
Next steps:
Intstead of "kNN prediction score = 41.6666666667%
Prediction accuracy appears rather poor. Observe other measures of performance." -- Make sure you add in benchmarking so you have an idea what "good" is. The long error output you're getting in your notebook is because you need to call np.ravel(y) to force your y to be the right dimension.
Clarity of research problem:
I did not see updated slides at time of review. I think the research question needs a bit of tweaking for clarity as it is now.
Code:
Your work process was easy to follow. I think your notebook should be condensed for easy reading by anyone wanting to review your portfolio. Extremely long print statements / outputs in cells should be snipped. Glad you were able to leverage beautifulsoup. Make sure you dedicate sufficient space to highlight your custom sentiment function.
Next steps:
Intstead of "kNN prediction score = 41.6666666667% Prediction accuracy appears rather poor. Observe other measures of performance." -- Make sure you add in benchmarking so you have an idea what "good" is. The long error output you're getting in your notebook is because you need to call np.ravel(y) to force your y to be the right dimension.