Cleanlab/student-grades

1---2license: mit3task_categories:4- tabular-classification5language:6- en7tags:8- education9- data-centric-ai10- label-noise11- cleanlab12pretty_name: Student Grades Dataset13size_categories:14- n<1K15---16 17# Student Grades Dataset18 19## Dataset Description20 21This dataset contains student grade data used in the cleanlab tutorial: [Improving ML Performance via Data Curation with Train vs Test Splits](https://docs.cleanlab.ai/stable/tutorials/improving_ml_performance.html).22 23The task is to predict each student's final letter grade (A, B, C, D, F) based on their exam scores and notes.24 25### Dataset Summary26 27- **Total Examples**: ~750 (train + test)28- **Task**: Multi-class classification29- **Features**:30  - `exam_1`: Score on first exam (0-100)31  - `exam_2`: Score on second exam (0-100)32  - `exam_3`: Score on third exam (0-100)33  - `notes`: Categorical notes about student (e.g., "great participation +10", "cheated on exam, gets 0pts")34  - `stud_ID`: Unique student identifier35- **Label**: `noisy_letter_grade` - Letter grade (A, B, C, D, F)36 37### Dataset Structure38 39```python40from datasets import load_dataset41 42dataset = load_dataset("cleanlab/student-grades")43 44# Access splits45train_data = dataset["train"]46test_data = dataset["test"]47 48# Convert to pandas49import pandas as pd50df_train = train_data.to_pandas()51df_test = test_data.to_pandas()52```53 54### Data Splits55 56| Split | Examples |57|-------|----------|58| train | ~600     |59| test  | ~130     |60 61### Dataset Fields62 63- **stud_ID** (string): Unique student identifier64- **exam_1** (float): First exam score (0-100)65- **exam_2** (float): Second exam score (0-100)66- **exam_3** (float): Third exam score (0-100)67- **notes** (string): Categorical notes about the student68- **noisy_letter_grade** (string): Final letter grade (A, B, C, D, F) - may contain label errors69 70## Dataset Creation71 72This dataset was created for educational purposes to demonstrate data-centric AI techniques using cleanlab. The data intentionally contains:73- **Label noise**: Some grades may be incorrectly labeled74- **Near duplicates**: Some examples are very similar or exact duplicates75- **Outliers**: Unusual data points that don't fit the distribution76 77These issues are introduced to help users learn how to detect and handle common data quality problems using cleanlab.78 79## Uses80 81### Primary Use Case82 83This dataset is designed for:841. Learning data-centric AI techniques852. Demonstrating cleanlab's capabilities for detecting label errors, outliers, and near duplicates863. Teaching proper train/test data curation workflows87 88### Example Usage89 90```python91from datasets import load_dataset92from cleanlab import Datalab93 94# Load dataset95dataset = load_dataset("cleanlab/student-grades")96df_train = dataset["train"].to_pandas()97 98# Use cleanlab to detect issues99lab = Datalab(data=df_train, label_name="noisy_letter_grade", task="classification")100lab.find_issues()101lab.report()102```103 104## Tutorial105 106For a complete tutorial using this dataset, see:107[Improving ML Performance via Data Curation with Train vs Test Splits](https://docs.cleanlab.ai/stable/tutorials/improving_ml_performance.html)108 109## Licensing Information110 111MIT License112 113## Citation114 115If you use this dataset in your research, please cite the cleanlab library:116 117```bibtex118@software{cleanlab,119  author = {Northcutt, Curtis G. and Athalye, Anish and Mueller, Jonas},120  title = {cleanlab},121  year = {2021},122  url = {https://github.com/cleanlab/cleanlab},123}124```125 126## Contact127 128- **Maintainers**: Cleanlab Team129- **Repository**: https://github.com/cleanlab/cleanlab130- **Documentation**: https://docs.cleanlab.ai131- **Issues**: https://github.com/cleanlab/cleanlab/issues132