0
$\begingroup$

I have an excel sheet filled with my own personal appreciations of movies I've watched, and I want to use it to train an AI model so that it can predict if I'll like a specific movie or not, based on the ones I've already seen.

My data is formatted as following (just a sample, the spreadsheet is filled with hundreds of movies):

enter image description here

And I would like to use all the columns to train my model. Because I am going to say if I liked the movie or not, I know it will be Supervised Learning. I already cleaned the data so there's no blank or missing data, but I do not know how to train my model using every column.

If required, I can be more specific on something, just ask and I'll edit the post.

$\endgroup$
  • $\begingroup$ Create an embedding of 'description' field, using word embeddings or perhaps contextualized embeddings from a language model. The 'director' field you could either one-hot encode, or create embeddings, since there might be similarities between directors in a vector space, those you could train from scratch using word2vec. The rating field you could use as it is by converting it to an integer. Then you could concatenate all the vectors and train a model, or train 3 different models and ensemble them :) The 'Movie Name' field does prob not contain any signal. Good luck :) $\endgroup$ – Isbister Oct 7 at 21:00
0
$\begingroup$

You will need to convert that to something that a neural network can understand.

  • Movie Name

Is useless. At least you want to judge a movie by its name.

  • Description

You will need to perform a Tokenization. Grab the x's more common words and convert them on an array.
I recommend you to see these videos from TensorFlow.

https://www.youtube.com/watch?v=fNxaJsNG3-s&t=18s

There you can find the Google Colab Links for the job.

  • Director

There are two options. A One Hot Array or an integer for every director.
One Hot Array may be better if you don't have an order of similarity of the directors. But will increase the size of the inputs.
If you have an order of similarity of the directors an integer for each director will work fine.

  • Rating

Nothing to do here. Is ready to go.

You can perform this work directly on excel, but tensorflow has great tools for it. It’s hard to run the model after you have trained it on excel.
If you are to comfortable on excel or you can install new software on your computer, I’ve made a backpropagation algorithm that runs on excel and give you a formula to paste in a module.

https://github.com/TorrensJoaquin/Multivariant-Nonlinear-Regression-in-VBA-Small-Neural-Network

| improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.