| 2026-02-16 02:26 | Added from scratch directory | Andrew Laack | 1 | +149 | -0 |
| 2024-11-15 02:16 | revised mat mul | Andrew | 1 | +32 | -6 |
| 2024-11-15 02:00 | Wrote code for matmul, dp vector matrix, and dp vector vector. | Andrew | 4 | +695 | -0 |
| 2024-11-08 01:26 | NN sorting seems... possibly a valid avenue of research. | Andrew | 2 | +391 | -1 |
| 2024-10-31 00:53 | Added prediction, this is very slow... | Andrew | 1 | +6782 | -135 |
| 2024-10-29 22:42 | Updated splitting logic to check between samples, fixed issue with duplicate samples different classifications, updated graph printing, fixed gini impurity calculation (is not weighted not summed) | Andrew | 1 | +186 | -252 |
| 2024-10-29 03:45 | did stuff | Andrew | 1 | +30727 | -0 |
| 2024-10-28 20:58 | Wrote basic code to do general decision tree related stuff. | Andrew | 1 | +463 | -21 |
| 2024-10-28 18:56 | Createde decision tree graphic | Andrew | 1 | +53 | -65 |
| 2024-10-28 05:13 | did some stuff with visualization and other stuff. | Andrew | 5 | +10169 | -43 |
| 2024-10-25 23:13 | Moved things | Andrew | 4 | +8192 | -348 |
| 2024-10-25 01:57 | finished some stuff | Andrew | 3 | +674 | -282 |
| 2024-10-24 22:52 | Worked on some stuff | Andrew | 3 | +663 | -134 |
| 2024-10-23 22:09 | Did stuff in stream. | Andrew | 2 | +1693 | -0 |
| 2024-08-08 23:11 | Wanted to mess with one hot encoder. | Andrew | 1 | +517 | -0 |
| 2024-08-08 21:15 | Stupid dataset. | Andrew | 1 | +1668 | -0 |
| 2024-07-17 23:03 | added polynomial feature logistic regression. | Andrew | 1 | +47 | -0 |
| 2024-07-17 22:53 | Diabetes prediction. | Andrew | 1 | +4740 | -0 |
| 2024-07-16 16:16 | Updated metrics for model eval. | Andrew | 1 | +641416 | -641399 |
| 2024-07-15 12:53 | Trained autoencoder and subsequent model to predict cc anomalies. | Andrew | 1 | +857807 | -0 |
| 2024-07-13 00:18 | Did stuff | Andrew | 2 | +125148 | -6 |
| 2024-07-11 22:16 | playing with compression | Andrew | 2 | +563682 | -0 |
| 2024-07-11 12:15 | Did stuff | Andrew | 3 | +10137 | -5 |
| 2024-07-09 21:15 | Did stuff | Andrew | 3 | +218592 | -307 |
| 2024-07-09 19:08 | plotly testing | Andrew | 1 | +1225 | -0 |
| 2024-07-08 04:27 | Did stuff | Andrew | 1 | +643 | -0 |
| 2024-07-01 23:20 | implemented data augmentation. | Andrew | 1 | +209 | -0 |
| 2024-07-01 20:59 | Created autoencoder for mnist. | Andrew | 2 | +245 | -0 |
| 2024-07-01 18:21 | 92% on mnist fashion with cnn. | Andrew | 2 | +368 | -0 |
| 2024-07-01 13:14 | Did some image and text manipulation stuff. | Andrew | 3 | +180 | -29 |
| 2024-07-01 03:29 | Created model. | Andrew | 1 | +2111 | -37 |
| 2024-06-30 16:52 | Did biol stuff and created new neural network for token prediction with word tokens instead of characters. | Andrew | 2 | +382 | -0 |
| 2024-06-29 18:14 | MN NB | Andrew | 2 | +1324 | -33 |
| 2024-06-28 20:03 | stuff | Andrew | 2 | +527 | -0 |
| 2024-06-27 22:04 | note | Andrew | 1 | +9 | -1 |
| 2024-06-27 22:03 | Used DNN to generate text based on shakespeare | Andrew | 1 | +2909 | -0 |
| 2024-06-27 15:42 | Predict cheating for students based on scores. | Andrew | 1 | +1253 | -0 |
| 2024-06-27 04:00 | Created some more models for student performance | Andrew | 4 | +2249 | -59 |
| 2024-06-26 21:57 | Used log reg and NN for recidivism calculation. They were both at 88% accuracy. Realistically, this is probably the best that is possible. While I could engineer features, I pretty much one hot encoded everything so there isn't much more to be found. | Andrew | 2 | +216 | -59 |
| 2024-06-26 21:43 | Fixed graph. Did some other stuff as well. | Andrew | 3 | +755 | -474 |
| 2024-06-26 19:20 | It all works now!!! | Andrew | 1 | +418 | -173 |
| 2024-06-26 19:05 | It's alive | Andrew | 1 | +292 | -58 |
| 2024-06-26 16:50 | Wrote the code to do a correct split using gini impurity. Next, I will write the code to create the entire tree. I also need to think of a way to compress the representation for prediction. | Andrew | 1 | +247 | -0 |
| 2024-06-25 22:31 | Created Neural Network for spam filtering achieving 86 accuracy. | Andrew | 1 | +342 | -0 |
| 2024-06-25 21:39 | Naive Bayes Custom Spam | Andrew | 1 | +744 | -0 |
| 2024-06-25 19:21 | Created CNN for mnist. | Andrew | 1 | +405 | -0 |
| 2024-06-25 02:13 | Did diagramming and some stuff | Andrew | 4 | +352 | -0 |
| 2024-06-14 20:04 | v | Andrew | 1 | +73 | -61 |
| 2024-06-11 16:05 | Completed mnist using NNs. The first time I was being really stupid and used regression and rounding to find the nearest int (stupid) while the second time I did it right with onehot encoding and softmax functions. | Andrew | 3 | +929 | -0 |
| 2024-06-11 12:04 | Started working with keras for deep learning. | Andrew | 3 | +799 | -0 |
| 2024-06-11 10:36 | Moved around files and began working with MLPs for regression and classification. I also worked through some clustering with DBSCAN (neighbors) and kmeans. | Andrew | 9 | +2376 | -0 |
| 2024-06-10 16:28 | Updated parameters for random forest and I used PCA projection to decrease the total number of dimensions needed to 23 from 784. | Andrew | 1 | +522 | -0 |
| 2024-06-10 14:56 | Completed notes and implementation of gradient boosting and adaboosting for ensembles. | Andrew | 2 | +2963 | -0 |
| 2024-06-10 01:01 | added voting classifier using soft voting and also added out of bag prediction with bagging decision tree ensemble (random forest). | Andrew | 1 | +1537 | -0 |
| 2024-06-07 06:04 | Worked on some data changing through pca, decision trees, and more decision trees with regression and classification tasks. Regression is weird with them and seems suboptimal in most cases where linear regression or lin reg with polynomial features would be better. | Andrew | 3 | +1299 | -0 |
| 2024-06-07 04:54 | some stuff | Andrew | 6 | +3086 | -51 |
| 2024-06-06 01:15 | Wrote code to do softmax regression, logistic regression, stochastic gd, and gd for polynomial regression. | Andrew | 4 | +2120 | -0 |
| 2024-06-05 03:47 | Finished working on classificaiton general stuff. I also wrote a custom implementation of linear regression using the closed-form method (formula), and the gradient descent way as well. | Andrew | 3 | +2756 | -56 |
| 2024-06-04 03:17 | Ending the day | Andrew | 1 | +41 | -0 |
| 2024-06-04 02:56 | Completed regression problem for housing using random forest. | Andrew | 5 | +1603 | -32 |
| 2024-05-30 02:23 | Worked on transforms to change distributions (log for right/left skew etc.) | Andrew | 2 | +381 | -25 |
| 2024-05-25 21:15 | updated | Andrew | 1 | +44 | -40 |
| 2024-05-25 21:15 | implemented standardization algorithm for pandas data frame. | Andrew | 2 | +144 | -44 |
| 2024-05-25 00:06 | Completed some stuff | Andrew | 3 | +343 | -280 |
| 2024-05-23 15:35 | Graphed sigmoid function | Andrew | 2 | +357 | -539 |
| 2024-05-23 02:47 | Created regression model with mapping for ocean distance from string to integer, rooms per house, bedroom ratio, and people per house. | Andrew | 1 | +651 | -56 |
| 2024-05-22 23:44 | Plotted scatter matrix | Andrew | 1 | +153 | -10 |
| 2024-05-18 01:43 | Completed implementation of person's correlation coefficient calculation. | Andrew | 1 | +93 | -6 |
| 2024-05-17 18:13 | Started work on correlation coefficient implementation and continued work on good lin reg for housing. | Andrew | 3 | +519 | -70 |
| 2024-05-16 01:37 | added start of book lin reg | Andrew | 1 | +151 | -0 |
| 2024-05-15 15:18 | Implemented linear regression on a dataset with many features to do univariate inference for house prices. | Andrew | 1 | +742 | -43 |
| 2024-05-15 02:11 | Added data and git ignore | Andrew | 5 | +171 | -124 |
| 2024-05-14 21:58 | Created implementation of RMSE/RMSD using python and arbitrary data. | Andrew | 3 | +125 | -7 |
| 2024-05-10 03:17 | changes | Andrew | 1 | +3 | -11 |
| 2024-05-10 03:17 | Working linear regression by hand. | Andrew | 7 | +255 | -0 |
| 2024-04-18 23:22 | Added learning rate parameter and improved output formatting | Andrew | 1 | +13 | -7 |
| 2024-04-18 23:08 | Improved clarity | Andrew | 1 | +45 | -0 |
| 2024-04-18 22:48 | rename file | Andrew | 2 | +46 | -47 |
| 2024-04-18 22:46 | Added code for gradient descent from notes | Andrew | 1 | +47 | -0 |
