Update README.md
Browse files
README.md
CHANGED
|
@@ -41,4 +41,37 @@ configs:
|
|
| 41 |
path: data/original-*
|
| 42 |
- split: augmented
|
| 43 |
path: data/augmented-*
|
|
|
|
|
|
|
| 44 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
path: data/original-*
|
| 42 |
- split: augmented
|
| 43 |
path: data/augmented-*
|
| 44 |
+
license: mit
|
| 45 |
+
pretty_name: NFL 2024 Receiving Yard Leaders
|
| 46 |
---
|
| 47 |
+
|
| 48 |
+
## Dataset Description
|
| 49 |
+
This dataset contains statistics for the **NFL 2024 receiving yard leaders**, with both original measurements and augmented samples. Each row corresponds to a player, with the target variable being **TD (touchdowns scored)**.
|
| 50 |
+
|
| 51 |
+
- **Original split:** 30 manually collected samples with real player names.
|
| 52 |
+
- **Augmented split:** 300 synthetic samples with the player name replaced by placeholder values like `Synthetic_Player_1`, `Synthetic_Player_2`, etc., created using data augmentation techniques.
|
| 53 |
+
|
| 54 |
+
## Features
|
| 55 |
+
- `player` (string): Player name, or placeholder for augmented data
|
| 56 |
+
- `Yds` (int32): Total receiving yards
|
| 57 |
+
- `TD` (int32): Total touchdowns scored (target variable)
|
| 58 |
+
- `20+` (int32): Receptions of 20+ yards
|
| 59 |
+
- `40+` (int32): Receptions of 40+ yards
|
| 60 |
+
- `LNG` (int32): Longest reception
|
| 61 |
+
- `Rec` (int32): Total receptions
|
| 62 |
+
- `1st` (int32): First down receptions
|
| 63 |
+
- `1st%` (float32): First down reception percentage
|
| 64 |
+
- `Rec FUM` (int32): Fumbles on receptions
|
| 65 |
+
- `Rec YAC/R` (float32): Yards after catch per reception
|
| 66 |
+
- `Tgts` (int32): Total targets
|
| 67 |
+
|
| 68 |
+
## Intended Use
|
| 69 |
+
This dataset is suitable for **regression or supervised learning tasks**, such as predicting touchdowns based on receiving stats. It is also intended for **educational purposes**, including exploring feature importance, synthetic data generation, and supervised learning workflows.
|
| 70 |
+
|
| 71 |
+
## Limitations
|
| 72 |
+
- Original split is small (30 samples) and represents only top-performing players.
|
| 73 |
+
- Augmented split uses synthetic player names and may not fully represent real-world distributions.
|
| 74 |
+
|
| 75 |
+
## Citation
|
| 76 |
+
If you use this dataset, please cite:
|
| 77 |
+
"Collected NFL 2024 Receiving Yard Leaders Dataset, augmented for educational purposes in a machine learning coursework project."
|