Uppsala University

Mini project submission and leaderboard

Instructions and files

Data: songs_to_classify.csv, training_data.csv, readme.txt
Getting-started code example: knn-example.py

Leader board

PositionGroupPerformanceComment (by the group)Time stamp
1 Group 4 93% 2018-09-21 18:22:23
1 Group 7 93% XGB 2018-09-22 15:20:35
1 Group 4 93% fgnpxvat 2018-09-26 09:36:56
1 Group 9 93% XGB 2018-09-27 22:55:47
1 Group 48 93% Gradient Boosting 2018-10-07 15:57:08
2 Group 5 92% RF 2018-09-21 18:27:23
2 Group 4 92% fgnpxvat2 2018-09-24 23:10:36
2 Group 21 92% 55b91ce802559bbc2264eb517308d515a77da29e 2018-09-25 15:46:17
2 Group 1 92% Vanilla Random Forest 2018-09-26 15:48:28
2 Group 49 92% GBT 2018-09-26 22:16:20
2 Group 11 92% GDB + GridSearch 2018-09-27 14:11:59
2 Group 12 92% Sk RF 2018-09-27 16:00:28
2 Group 5 92% rnd1 2018-09-28 19:54:09
2 Group 1 92% GDB 2018-09-29 19:39:01
2 Group 37 92% First_attempt_m 2018-10-01 22:13:13
2 Group 37 92% Second_attempt 2018-10-02 22:00:08
2 Group 22 92% Boosting 2018-10-02 23:09:31
2 Group 8 92% RF 2018-10-07 23:15:16
2 Group 14 92% simple random forest 2018-10-10 22:58:44
3 Group 47 91% Random Random Forest 2018-09-20 13:31:16
3 Group 6 91% 1stTry 2018-09-23 20:25:13
3 Group 47 91% Gradient Boosting 2018-09-24 15:26:20
3 Group 10 91% 2018-09-24 19:08:25
3 Group 21 91% 34e18bf9722ebcd85008dc83ef54e68bcb24f317 2018-09-24 19:26:02
3 Group 9 91% 1st test aiming for 85.2% 2018-09-25 17:05:32
3 Group 12 91% Sk_Test 2018-09-26 16:25:08
3 Group 50 91% Random forest 2018-09-28 11:28:52
3 Group 36 91% rf 2018-10-01 11:10:49
3 Group 35 91% Rainforestt 2018-10-02 16:11:22
3 Group 21 91% RF 2018-10-02 18:50:22
3 Group 34 91% double rainbow 2018-10-08 23:34:06
4 Group 4 90% 2018-09-22 23:48:09
5 Group 1 89% The magic 7 2018-09-27 10:56:50
6 Group 9 88% 2018-09-26 13:05:11
6 Group 1 88% Magic 9 2018-09-28 08:53:53
6 Group 33 88% Voting 2018-10-03 17:34:49
7 Group 4 87% fgnpxvat 2018-09-23 23:36:54
7 Group 1 87% jr nyfb qvq fbzr fgnpxvat 2018-10-02 11:37:56
7 Group 7 87% RF 2018-10-06 23:39:09
8 Group 36 86% soft 2018-09-29 18:06:48
8 Group 33 86% MLP, k=13, (50,2) 2018-10-07 23:44:41
9 Group 13 85% SVM+ Gaussian Kernel 2018-09-25 13:00:50
9 Group 49 85% PCA+SVC 2018-09-27 10:47:55
9 Group 19 85% XGB 2018-09-27 13:45:15
9 Group 36 85% Initial 2018-09-28 10:50:46
10 Group 35 84% Rainforestt4 2018-10-03 10:45:54
10 Group 48 84% Random Forest 2018-10-08 23:01:42
11 Group 35 83% RainforestMLP 2018-10-09 20:24:28
12 Group 36 82% Hard 2018-09-30 18:58:23
12 Group 33 82% knn 1511 2018-10-01 23:40:34
13 Group 34 81% naive 2018-09-25 14:54:23
13 Group 1 81% First test of NN 2018-09-25 15:13:47
13 Group 49 81% PCA+SVC_2nd 2018-10-01 14:55:51
13 Group 59 81% NN 2018-11-16 09:25:35
14 Group 13 79% Logistic loss 2018-09-26 15:07:45
14 Group 33 79% LDA 2018-10-08 21:29:05
15 Group 33 78% Gradient Boosting 2018-09-21 15:38:21
15 Group 10 78% Random Random Forest 2018-09-23 23:59:52
16 Group 21 77% TEST 2018-09-23 23:32:43
16 Group 19 77% kNN, k=5, 2 features are categorized, distance weight 2018-09-28 11:40:33
17 Group 8 75% 2018-10-02 20:02:43
18 Group 8 70% RF 2018-10-08 00:00:18
19 Group 33 61% GPTS, k=10, n=5 2018-10-06 23:53:19
20 Group 59 60% Neural Network, Martin Ka 2018-11-14 14:43:52
21 Group 59 59% Martin Ka, ANN 2018-11-15 09:58:23
22 BoB 58% LIKE ALL 2018-09-04 09:34:03
23 Group 33 47% Gradient Boosting2 2018-09-28 18:39:15
Only one submission per group and day is shown. If several submissions are submitted from the same group in the same day, only the latest one is shown. Submissions made today will be shown from tomorrow.

Submit new contribution

Note! You should carefully use the training data to select ONE (1) method which you want to test 'in production' by submitting its predictions to this page. In order to complete the project, it is thus necessary to submit only ONE (1) contribution during the entire project! If you want to improve your solution and compete with the other groups, we do, however, allow you to submit up to one new solution per day. The competition is just for fun, your score on the leader board will not be used in any way to determine your grade on the project. The grading is based solely on your implementation, conveyed understanding, and the quality of the written report.

Group password: (passwords will be distributed when the project starts)
Prediction: (0 = dislike, 1 = like. No other charachters than 0, 1 and space are accepted. Example: 0 1 0 1 1 0 0 1 0 )
Label: (If you want to give your submission some label. Will be shown in the leader board to everyone. Special characters, ä, å, ö, etc, are not allowed.)

Technical problems?

Contact Andreas Svensson

About the project

Originally developed by Andreas Svensson for the master course Statistical Machine Learning at Uppsala University.