Differences

This shows you the differences between two versions of the page.

--- cs401r_w2016:lab12 [2016/03/28 15:59]
admin
+++ cs401r_w2016:lab12 [2021/06/30 23:42] (current)
@@ Line 6: / Line 6: @@
 ====Deliverable:====
-For this lab, you will construct a movie recommendation engine, using a simple publicly available dataset.  For this lab, you will turn in two things:
+For this lab, you will construct a movie recommendation engine, using a simple publicly available dataset.  For this lab, you will turn in three things:
   - A notebook containing your code, but we will not run it.
-  - A set of predictions for a specific list of <user,movie> pairs.
+  - A set of predictions for a specific list of <user,movie> pairs, in a CSV file.
+  - A report discussing your approach, how well it worked (in terms of RMSE), and any visualizations or patterns you found in the data.  Markdown format, please!!
 We will run a small "competition" on your predictions: the three students with the best predictions will get 10% extra credit on this lab.
@@ Line 20: / Line 21: @@
 Your entry will be graded on the following elements:
-  * 25% Correct implementation of Metropolis Hastings inference
+  * 85% Project writeup
-  * 5% Correct calculation of gradients
+    * 30% Exploratory data analysis
-  * 45% Correct implementation of Hamiltonian MCMC
+    * 30% Description of technical approach
-  * 15% A small write-up comparing and contrasting MH, HMC, and the different proposal distributions
+    * 25% Analysis of performance of method
-  * 10% Final plot is tidy and legible
+  * 15% Submission of predictions csv file
   * 10% extra credit for the three top predictions
@@ Line 35: / Line 35: @@
 The training set you will use can be downloaded here:
-[[http://hatch.cs.byu.edu/courses/stat_ml/movie_training_data.tar.gz|Movie ratings training data]]
+[[https://www.dropbox.com/s/7r4rebn1ytugi4g/movie_training_data.tar.gz?dl=0|Movie ratings training data]]
-A complete description of the data can be found in the ''readme.txt'' file.  This dataset is richer than the Netflix competition dataset; for each movie, you also have a corresponding IMDB ID, some RottenTomatoes information, as well as a set of tags that users may have used when rating each movie.
+You will need to make predictions for a set of ''user,movie'' pairs.  These can be downloaded here:
+[[https://www.dropbox.com/s/2xvsrju9w3fqfon/predictions.dat?dl=0|Movie predictions data]]
+A complete description of the data can be found in the ''readme.txt'' file.  This dataset is richer than the Netflix competition dataset; for each movie, you also have a director and genre information, a corresponding IMDB ID, some RottenTomatoes information, as well as a set of tags that users may have used when rating each movie.
 You should start by looking at the ''user_ratedmovies_train.dat'' file.  It is a CSV file containing user,movie,timestamp tuples that form the core training data.  Everything else is auxiliary data that may or may not be useful.
+**Turning in your submissions**
+As part of this lab, you must submit a set of predictions.  You must provide predictions as a simple CSV file with two columns and 85,000 rows.  Each row has the form
+''testID,predicted_rating''
+The ''testID'' field uniquely identifies each ''user,movie'' prediction pair in the predictions set.
+**Evaluating your submissions**
+Performance of your prediction engine will be based on RMSE:
+$$ \mathrm{RMSE} = \sqrt{ \frac{1}{N} \sum_{i} (\mathrm{prediction_i} - \mathrm{truth_i})^2 } $$
+**Note: it is strongly encouraged that you first partition your dataset into a training and a validation set, to assess the generalization performance of your rating algorithm!**
+**Project writeup**
+Because you are being given full freedom in choosing your implementation strategy, you will not be graded on it (except to ensure that your implementation matches what you describe in your writeup!).  Instead, you will be graded solely on a writeup describing your implementation.
+This writeup must include three main sections:
+  - **A discussion of your exploration of the dataset**.
+    - Before you start coding, you should look at the data.  What does it include?  What patterns do you see?
+    - Any visualizations about the data you deem relevant
+  - **A clear, technical description of your approach.**  This section should include:
+    - Background on the approach
+    - Description of the model you use
+    - Description of the inference / training algorithm you use
+    - Description of how you partitioned your data into a test/training split
+  - **An analysis of how your approach worked on the dataset**
+    - What was your final RMSE on your private test/training split?
+    - Did you overfit?  How do you know?
+    - Was your first algorithm the one you ultimately used for your submission?  Why did you (or didn't you) iterate your design?
 ----
@@ Line 48: / Line 89: @@
 import seaborn
 import pandas
+import numpy as np
 ur = pandas.read_csv('user_ratedmovies_train.dat','\t')
 plt.hist( ur['rating'] )
+# create a test/train split
+all_inds = np.random.permutation( range(0,len(ur)) )
+test_inds = all_inds[0:85000]
+train_inds = all_inds[85000:len(ur)]
+ur_test = ur.iloc[ test_inds ]
+ur_train = ur.iloc[ train_inds ]
 </code>
+And Here is some code that writes out your prediction file that you will submit:
+<code python>
+import numpy as np
+import pandas as pd
+pred_array = pd.read_table('predictions.dat')
+test_ids = pred_array[["testID"]]
+pred_array.head()
+N = pred_array.shape[0]
+my_preds = np.zeros((N,1))
+for id in range(N): ### Prediction loop
+    predicted_rating = 3
+    my_preds[ id, 0 ] = predicted_rating ### This Predicts everything as 3
+sfile = open( 'predictions.csv', 'w' )
+sfile.write( '"testID","predicted_rating"\n' )
+for id in range( 0, N ):
+    sfile.write( '%d,%.2f\n' % (test_ids.iloc[id], my_preds[id] ) )
+sfile.close()
+</code>

BYU CS classes

User Tools

Site Tools

Differences

Page Tools