Santostang / box-office-prediction

Cornell ORIE 4741 Course Project: Machine Learning with Big Messy Data
6 stars 5 forks source link

Final Peer Review #13

Open austonli opened 6 years ago

austonli commented 6 years ago

This project aims to determine box office performance of different movies during the opening weekend. This can be potentially useful as marketing agencies and advertisers can leverage this data to adjust their campaigns and strategies.

Strengths I love the in depth explanations of each feature. This really helps the reader understand what type of data the project works with. They do a great job at introducing the reader to the source of the data and why they selected it.

The selected models performed well in general, and the model selection does a nice job detailing their choices and the types of errors they encountered.

Weaknesses The dataset is quite small, and the there doesn't seem to be much consideration or suggested improvements to account for this. It is extremely important to realize the shortcomings of such a dataset and quantitatively capture this potential error.

The grammar is also quite distracting. In one graph, the caption says "Prediction of Opening Gross on Test Set", and within the graph the word opening is spelled as "openning" four times. There shouldn't quite be anything as glaringly obvious as this type of mistake.

While the very few data visualizations are generally quite pertinent, there seems to be no real reason to have an entire page dedicated to tiny pairwise plots. The report only touches on these graphs and other plots on a surface level.

The conclusion at the end seems a little lackluster. This may be due to the fact that the impetus behind the project is not clearly defined. The practical implications detailed at the end are very cursory. The project does a commendable job at analyzing the data, but should do better at forming real world applications.

Closing Remarks While the above remarks may seem a little overly critical, I still think the project does a satisfactory job at answering their objective. Some key improvements in the above areas would take this project to a new level of refinement.

Santostang commented 6 years ago

Thank you for you comment, I totally understand it's small, and I think we have mentioned it in the first and last part of the report. I felt this project is really interesting so I will scrape more data in the winter holiday although it really takes a lot of time to scrape and clean.

Thank you for pointing out the grammar issues. However, I don't quite understand your points of conclusion part, could you show me an example of how to form real world applications?