Dark Mode
Adult Autism Spectrum Predictor
Data Science and Analytics
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
Data details the results of the Autism Spectrum Quotient (AQ) 10 item screening tool, designed to assist in predicting Autism Spectrum Disorder (ASD) status in individuals aged 18 and over. This product includes various demographic and historical factors, such as age, gender, ethnicity, and family history of autism, alongside the raw and final classification scores. It is structured for machine learning applications, specifically binary classification, where the goal is to predict the 'Class/ASD' outcome (Yes or No).
Columns
The dataset contains 21 distinct features pertaining to patient characteristics and screening results:
- ID: Unique identifier for the patient.
- A1_Score to A10_Score: Individual scores derived from the AQ 10 item screening tool.
- age: The patient's age quantified in years.
- gender: The declared gender of the patient.
- ethnicity: The patient’s specific ethnic background.
- jaundice: A boolean indicator noting whether the patient experienced jaundice at the time of birth.
- autism: A boolean indicator specifying if an immediate family member has received an autism diagnosis.
- contry_of_res: The patient’s country of residence.
- used_app_before: A boolean indicator showing if the patient had previously completed a screening test.
- result: The calculated total score derived from the AQ1-10 screening test (ranging from 0 to 10).
- age_desc: A categorical description of the patient's age group (uniform value: '18 and more').
- relation: Describes the relationship of the individual who completed the test (e.g., Self, Parent).
- Class/ASD: The target variable, indicating the classified result as 0 (No ASD) or 1 (Yes ASD).
Distribution
The data is provided in CSV format and is separated into three files:
train.csv, test.csv, and sample_submission.csv. The total file size is approximately 8 kB, although the primary screening file totals 58.58 kB. The dataset contains 704 total values across all records and features 21 columns. The data is static, with an expected update frequency of "Never."Usage
This screening data is highly valuable for developing and testing classification models aimed at predicting ASD outcomes based on demographic factors and AQ screening scores. Ideal applications include:
- Building machine learning models (e.g., decision trees, logistic regression) to classify individuals into ASD or non-ASD groups.
- Performing exploratory data analysis to identify significant correlations between demographic attributes (like ethnicity, family autism history, or jaundice) and screening results.
- Evaluating the predictive power of the Autism Spectrum Quotient (AQ) scores.
- Practising data analytics techniques, especially for beginners.
Coverage
The data covers 704 records of adult patients, categorised by the age description '18 and more'. The demographic scope includes diverse groups, with high representation from the White-European and Asian ethnicities. Patients reside in numerous countries, with 'United States' and 'United Arab Emirates' being the most frequently recorded countries of residence. Approximately 27% of the recorded classifications resulted in a positive ASD classification.
License
CC0: Public Domain
Who Can Use It
- Data Scientists and Machine Learning Engineers: For building, training, and validating predictive classification algorithms.
- Students and Beginners in Data Analytics: For learning basic data manipulation and model creation using a structured tabular dataset.
- Researchers in Psychology or Health Informatics: For studying the relationship between the AQ scores, demographic variables, and final ASD classification.
Dataset Name Suggestions
- ASD Screening Classification Data
- Adult Autism Spectrum Predictor
- AQ-10 Item Screening Results
- Jaundice and Family Autism History Dataset
Attributes
Original Data Source: Adult Autism Spectrum Predictor
QUALITY