File size: 20,291 Bytes
8176fea
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "998997dd",
   "metadata": {},
   "source": [
    "# Modeling NCAA Tournament Basketball games\n",
    "\n",
    "The thought process is to build a neural network that can predict a teams tournament <br>\n",
    "performance on a per game basis. Then we can use these predicted metrics to run a monte carlo <br>\n",
    "style simulation and select whichever team is most likley to win. <br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "f0ec30d9",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "\n",
    "import pandas as pd\n",
    "import os\n",
    "\n",
    "\n",
    "# check to make sure if there are any gpu's available for faster training\n",
    "def get_device() -> str:\n",
    "    if torch.cuda.is_available():\n",
    "        return \"cuda\"\n",
    "    if torch.backends.mps.is_available():\n",
    "        return \"mps\" \n",
    "    return \"cpu\"\n",
    "\n",
    "# mps not working correctly on my m1 macbook air so just doing cpu for now\n",
    "# DEVICE = get_device()\n",
    "DEVICE = \"cpu\"\n",
    "\n",
    "# universal data directory for this project\n",
    "DATA_DIR = os.path.join(\"..\", \"data\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "b820f210",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 655 entries, 0 to 654\n",
      "Columns: 1068 entries, Unnamed: 0 to Seed\n",
      "dtypes: float64(672), int64(388), object(8)\n",
      "memory usage: 5.3+ MB\n"
     ]
    }
   ],
   "source": [
    "all_games_df = pd.read_csv(os.path.join(DATA_DIR, \"MDetailedAggregatedGames.csv\"))\n",
    "all_games_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "02ebc500",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>Season</th>\n",
       "      <th>DayNum</th>\n",
       "      <th>WTeamID</th>\n",
       "      <th>WScore</th>\n",
       "      <th>LTeamID</th>\n",
       "      <th>LScore</th>\n",
       "      <th>WLoc</th>\n",
       "      <th>NumOT</th>\n",
       "      <th>WFGM</th>\n",
       "      <th>...</th>\n",
       "      <th>tourney_DR_max</th>\n",
       "      <th>tourney_DR_mean</th>\n",
       "      <th>tourney_DR_median</th>\n",
       "      <th>tourney_DR_std</th>\n",
       "      <th>tourney_DR_sum</th>\n",
       "      <th>ConfAbbrev</th>\n",
       "      <th>TeamName</th>\n",
       "      <th>FirstD1Season</th>\n",
       "      <th>LastD1Season</th>\n",
       "      <th>Seed</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>2003</td>\n",
       "      <td>40</td>\n",
       "      <td>1266</td>\n",
       "      <td>63</td>\n",
       "      <td>1458</td>\n",
       "      <td>54</td>\n",
       "      <td>H</td>\n",
       "      <td>0</td>\n",
       "      <td>24</td>\n",
       "      <td>...</td>\n",
       "      <td>21.666667</td>\n",
       "      <td>21.666667</td>\n",
       "      <td>21.666667</td>\n",
       "      <td>21.666667</td>\n",
       "      <td>21.666667</td>\n",
       "      <td>big_ten</td>\n",
       "      <td>Wisconsin</td>\n",
       "      <td>1985</td>\n",
       "      <td>2024</td>\n",
       "      <td>Y05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>2003</td>\n",
       "      <td>97</td>\n",
       "      <td>1266</td>\n",
       "      <td>68</td>\n",
       "      <td>1448</td>\n",
       "      <td>61</td>\n",
       "      <td>H</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>...</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>acc</td>\n",
       "      <td>Wake Forest</td>\n",
       "      <td>1985</td>\n",
       "      <td>2024</td>\n",
       "      <td>W02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9</td>\n",
       "      <td>2003</td>\n",
       "      <td>115</td>\n",
       "      <td>1266</td>\n",
       "      <td>78</td>\n",
       "      <td>1257</td>\n",
       "      <td>73</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>26</td>\n",
       "      <td>...</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>cusa</td>\n",
       "      <td>Louisville</td>\n",
       "      <td>1985</td>\n",
       "      <td>2024</td>\n",
       "      <td>W04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12</td>\n",
       "      <td>2003</td>\n",
       "      <td>138</td>\n",
       "      <td>1266</td>\n",
       "      <td>101</td>\n",
       "      <td>1281</td>\n",
       "      <td>92</td>\n",
       "      <td>N</td>\n",
       "      <td>1</td>\n",
       "      <td>35</td>\n",
       "      <td>...</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>big_twelve</td>\n",
       "      <td>Missouri</td>\n",
       "      <td>1985</td>\n",
       "      <td>2024</td>\n",
       "      <td>Y06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19</td>\n",
       "      <td>2003</td>\n",
       "      <td>143</td>\n",
       "      <td>1266</td>\n",
       "      <td>77</td>\n",
       "      <td>1338</td>\n",
       "      <td>74</td>\n",
       "      <td>N</td>\n",
       "      <td>0</td>\n",
       "      <td>28</td>\n",
       "      <td>...</td>\n",
       "      <td>21.333333</td>\n",
       "      <td>21.333333</td>\n",
       "      <td>21.333333</td>\n",
       "      <td>21.333333</td>\n",
       "      <td>21.333333</td>\n",
       "      <td>big_east</td>\n",
       "      <td>Pittsburgh</td>\n",
       "      <td>1985</td>\n",
       "      <td>2024</td>\n",
       "      <td>Y02</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 1068 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  Season  DayNum  WTeamID  WScore  LTeamID  LScore WLoc  NumOT  \\\n",
       "0           0    2003      40     1266      63     1458      54    H      0   \n",
       "1           5    2003      97     1266      68     1448      61    H      0   \n",
       "2           9    2003     115     1266      78     1257      73    A      0   \n",
       "3          12    2003     138     1266     101     1281      92    N      1   \n",
       "4          19    2003     143     1266      77     1338      74    N      0   \n",
       "\n",
       "   WFGM  ...  tourney_DR_max  tourney_DR_mean  tourney_DR_median  \\\n",
       "0    24  ...       21.666667        21.666667          21.666667   \n",
       "1    21  ...       26.000000        26.000000          26.000000   \n",
       "2    26  ...       24.000000        24.000000          24.000000   \n",
       "3    35  ...       26.000000        26.000000          26.000000   \n",
       "4    28  ...       21.333333        21.333333          21.333333   \n",
       "\n",
       "   tourney_DR_std  tourney_DR_sum  ConfAbbrev     TeamName  FirstD1Season  \\\n",
       "0       21.666667       21.666667     big_ten    Wisconsin           1985   \n",
       "1       26.000000       26.000000         acc  Wake Forest           1985   \n",
       "2       24.000000       24.000000        cusa   Louisville           1985   \n",
       "3       26.000000       26.000000  big_twelve     Missouri           1985   \n",
       "4       21.333333       21.333333    big_east   Pittsburgh           1985   \n",
       "\n",
       "   LastD1Season  Seed  \n",
       "0          2024   Y05  \n",
       "1          2024   W02  \n",
       "2          2024   W04  \n",
       "3          2024   Y06  \n",
       "4          2024   Y02  \n",
       "\n",
       "[5 rows x 1068 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_games_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58e4fee8",
   "metadata": {},
   "source": [
    "# Feature Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1251726e",
   "metadata": {},
   "outputs": [],
   "source": [
    "target_df = all_games_df[[\"tourney_Score_mean\", \"tourney_Score_std\", \"tourney_Score_max\", \"tourney_Score_min\"]]\n",
    "\n",
    "features_df = all_games_df[[col for col in all_games_df if col.startswith(\"reg\") and \"_W\" not in col and \"_L\" not in col and \"sum\" not in col]]\n",
    "# features_df = features_df.select_dtypes(include=\"number\")\n",
    "\n",
    "# split data into training and testing data sets\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    features_df.astype(float),\n",
    "    target_df.astype(float),\n",
    "    train_size=0.8,\n",
    "    random_state=8,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "28478189",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 524 entries, 5 to 451\n",
      "Data columns (total 71 columns):\n",
      " #   Column            Non-Null Count  Dtype  \n",
      "---  ------            --------------  -----  \n",
      " 0   reg_Games         524 non-null    float64\n",
      " 1   reg_Score_min     524 non-null    float64\n",
      " 2   reg_Score_max     524 non-null    float64\n",
      " 3   reg_Score_mean    524 non-null    float64\n",
      " 4   reg_Score_median  524 non-null    float64\n",
      " 5   reg_Score_std     524 non-null    float64\n",
      " 6   reg_FGM_min       524 non-null    float64\n",
      " 7   reg_FGM_max       524 non-null    float64\n",
      " 8   reg_FGM_mean      524 non-null    float64\n",
      " 9   reg_FGM_median    524 non-null    float64\n",
      " 10  reg_FGM_std       524 non-null    float64\n",
      " 11  reg_FGA_min       524 non-null    float64\n",
      " 12  reg_FGA_max       524 non-null    float64\n",
      " 13  reg_FGA_mean      524 non-null    float64\n",
      " 14  reg_FGA_median    524 non-null    float64\n",
      " 15  reg_FGA_std       524 non-null    float64\n",
      " 16  reg_FTM_min       524 non-null    float64\n",
      " 17  reg_FTM_max       524 non-null    float64\n",
      " 18  reg_FTM_mean      524 non-null    float64\n",
      " 19  reg_FTM_median    524 non-null    float64\n",
      " 20  reg_FTM_std       524 non-null    float64\n",
      " 21  reg_FTA_min       524 non-null    float64\n",
      " 22  reg_FTA_max       524 non-null    float64\n",
      " 23  reg_FTA_mean      524 non-null    float64\n",
      " 24  reg_FTA_median    524 non-null    float64\n",
      " 25  reg_FTA_std       524 non-null    float64\n",
      " 26  reg_Ast_min       524 non-null    float64\n",
      " 27  reg_Ast_max       524 non-null    float64\n",
      " 28  reg_Ast_mean      524 non-null    float64\n",
      " 29  reg_Ast_median    524 non-null    float64\n",
      " 30  reg_Ast_std       524 non-null    float64\n",
      " 31  reg_Blk_min       524 non-null    float64\n",
      " 32  reg_Blk_max       524 non-null    float64\n",
      " 33  reg_Blk_mean      524 non-null    float64\n",
      " 34  reg_Blk_median    524 non-null    float64\n",
      " 35  reg_Blk_std       524 non-null    float64\n",
      " 36  reg_PF_min        524 non-null    float64\n",
      " 37  reg_PF_max        524 non-null    float64\n",
      " 38  reg_PF_mean       524 non-null    float64\n",
      " 39  reg_PF_median     524 non-null    float64\n",
      " 40  reg_PF_std        524 non-null    float64\n",
      " 41  reg_Stl_min       524 non-null    float64\n",
      " 42  reg_Stl_max       524 non-null    float64\n",
      " 43  reg_Stl_mean      524 non-null    float64\n",
      " 44  reg_Stl_median    524 non-null    float64\n",
      " 45  reg_Stl_std       524 non-null    float64\n",
      " 46  reg_TO_min        524 non-null    float64\n",
      " 47  reg_TO_max        524 non-null    float64\n",
      " 48  reg_TO_mean       524 non-null    float64\n",
      " 49  reg_TO_median     524 non-null    float64\n",
      " 50  reg_TO_std        524 non-null    float64\n",
      " 51  reg_FGM3_min      524 non-null    float64\n",
      " 52  reg_FGM3_max      524 non-null    float64\n",
      " 53  reg_FGM3_mean     524 non-null    float64\n",
      " 54  reg_FGM3_median   524 non-null    float64\n",
      " 55  reg_FGM3_std      524 non-null    float64\n",
      " 56  reg_FGA3_min      524 non-null    float64\n",
      " 57  reg_FGA3_max      524 non-null    float64\n",
      " 58  reg_FGA3_mean     524 non-null    float64\n",
      " 59  reg_FGA3_median   524 non-null    float64\n",
      " 60  reg_FGA3_std      524 non-null    float64\n",
      " 61  reg_OR_min        524 non-null    float64\n",
      " 62  reg_OR_max        524 non-null    float64\n",
      " 63  reg_OR_mean       524 non-null    float64\n",
      " 64  reg_OR_median     524 non-null    float64\n",
      " 65  reg_OR_std        524 non-null    float64\n",
      " 66  reg_DR_min        524 non-null    float64\n",
      " 67  reg_DR_max        524 non-null    float64\n",
      " 68  reg_DR_mean       524 non-null    float64\n",
      " 69  reg_DR_median     524 non-null    float64\n",
      " 70  reg_DR_std        524 non-null    float64\n",
      "dtypes: float64(71)\n",
      "memory usage: 294.8 KB\n"
     ]
    }
   ],
   "source": [
    "X_train.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "04f4a0a6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 524 entries, 5 to 451\n",
      "Data columns (total 4 columns):\n",
      " #   Column              Non-Null Count  Dtype  \n",
      "---  ------              --------------  -----  \n",
      " 0   tourney_Score_mean  524 non-null    float64\n",
      " 1   tourney_Score_std   524 non-null    float64\n",
      " 2   tourney_Score_max   524 non-null    float64\n",
      " 3   tourney_Score_min   524 non-null    float64\n",
      "dtypes: float64(4)\n",
      "memory usage: 20.5 KB\n"
     ]
    }
   ],
   "source": [
    "y_train.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "40094cd0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# convert all datasets into tensors and register them \n",
    "# with the device (cuda, mps or cpu)\n",
    "X_trainT = torch.Tensor(\n",
    "    X_train.values,\n",
    ").float().to(DEVICE)\n",
    "\n",
    "X_testT = torch.Tensor(\n",
    "    X_test.values,\n",
    ").float().to(DEVICE)\n",
    "\n",
    "y_trainT = torch.Tensor(\n",
    "    y_train.values,\n",
    ").float().to(DEVICE)\n",
    "\n",
    "y_testT = torch.Tensor(\n",
    "    y_test.values,\n",
    ").float().to(DEVICE)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "20bceb9a",
   "metadata": {},
   "source": [
    "# Building Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "7b0573ee",
   "metadata": {},
   "outputs": [],
   "source": [
    "num_features = len(X_train.columns)\n",
    "\n",
    "class MadnessNN(nn.Module):\n",
    "    def __init__(self) -> None:\n",
    "        super().__init__()\n",
    "        self.input_layer = nn.Linear(num_features, 64)\n",
    "        self.activation_func = nn.ReLU()\n",
    "        self.layer1 = nn.Linear(64, 32)\n",
    "        self.layer2 = nn.Linear(32, 16)\n",
    "        self.layer3 = nn.Linear(16, 8)\n",
    "        self.output_layer = nn.Linear(8, 4)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.input_layer(x)\n",
    "        x = self.activation_func(x)\n",
    "        x = self.layer1(x)\n",
    "        x = self.activation_func(x)\n",
    "        x = self.layer2(x)\n",
    "        x = self.activation_func(x)\n",
    "        x = self.layer3(x)\n",
    "        x = self.activation_func(x)\n",
    "        x = self.output_layer(x)\n",
    "        x = self.activation_func(x)\n",
    "        return x\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "061e2b52",
   "metadata": {},
   "source": [
    "# Training Loop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "db035b9d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[500 / 5000] Loss = 40.454681396484375\n",
      "[1000 / 5000] Loss = 39.701454162597656\n",
      "[1500 / 5000] Loss = 39.055484771728516\n",
      "[2000 / 5000] Loss = 38.53948974609375\n",
      "[2500 / 5000] Loss = 38.149085998535156\n",
      "[3000 / 5000] Loss = 37.87413024902344\n",
      "[3500 / 5000] Loss = 37.6934928894043\n",
      "[4000 / 5000] Loss = 37.573673248291016\n",
      "[4500 / 5000] Loss = 37.48927307128906\n",
      "[5000 / 5000] Loss = 37.43183135986328\n"
     ]
    }
   ],
   "source": [
    "torch.manual_seed(1)\n",
    "\n",
    "model5000 = MadnessNN()\n",
    "optimizer = optim.Adam(lr=0.001, params=model5000.parameters())\n",
    "loss_fn = nn.MSELoss()\n",
    "epochs = 5000\n",
    "\n",
    "for epoch in range(1, epochs + 1):\n",
    "    pred = model5000(X_trainT)\n",
    "    loss = loss_fn(pred, y_trainT)\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "    optimizer.zero_grad()\n",
    "\n",
    "    if epoch % 500 == 0:\n",
    "        print(f\"[{epoch} / {epochs}] Loss = {loss}\") \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "b62fd19c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# save\n",
    "torch.save(model5000, os.path.join(\"models\", \"model5000.pth\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "17694dc7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE on testing data: 47.071144104003906\n"
     ]
    }
   ],
   "source": [
    "# evaluate\n",
    "model5000.eval()\n",
    "\n",
    "with torch.no_grad():\n",
    "    pred = model5000(X_testT)\n",
    "    loss = loss_fn(pred, y_testT)\n",
    "    print(f\"MSE on testing data: {loss}\")\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}