Predicting Risky Sexual Behavior Among College Students Through Machine Learning Approaches: Cross-sectional Analysis of Individual Data From 1264 Universities in 31 Provinces in China
Bibliographic record
Abstract
BACKGROUND: Risky sexual behavior (RSB), the most direct risk factor for sexually transmitted infections (STIs), is common among college students. Thus, identifying relevant risk factors and predicting RSB are important to intervene and prevent RSB among college students. OBJECTIVE: We aim to establish a predictive model for RSB among college students to facilitate timely intervention and the prevention of RSB to help limit STI contraction. METHODS: We included a total of 8794 heterosexual Chinese students who self-reported engaging in sexual intercourse from November 2019 to February 2020. We identified RSB among those students and attributed it to 4 dimensions: whether contraception was used, whether the contraceptive method was safe, whether students engaged in casual sex or sex with multiple partners, and integrated RSB (which combined the first 3 dimensions). Overall, 126 predictors were included in this study, including demographic characteristics, daily habits, physical and mental health, relationship status, sexual knowledge, sexual education, sexual attitude, and previous sexual experience. For each type of RSB, we compared 8 machine learning (ML) models: multiple logistic regression (MLR), naive Bayes (BYS), linear discriminant analysis (LDA), random forest (RF), gradient boosting machine (GBM), extreme gradient boosting (XGBoost), deep learning (DL), and the ensemble model. The optimal model for both RSB prediction and risk factor identification was selected based on a set of validation indicators. An MLR model was applied to investigate the association between RSB and identified risk factors through ML methods. RESULTS: In total, 5328 (60.59%) students were found to have previously engaged in RSB. Among them, 3682 (41.87%) did not use contraception every time they had sexual intercourse, 3602 (40.96%) had previously used an ineffective or unsafe contraceptive method, and 1157 (13.16%) had engaged in casual sex or sex with multiple partners. XGBoost achieved the optimal predictive performance on all 4 types of RSB, with the area under the receiver operator characteristic curve (AUROC) reaching 0.78, 0.72, 0.94, and 0.80 for contraceptive use, safe contraceptive method use, engagement in casual sex or with multiple partners, and integrated RSB, respectively. By ensuring the stability of various validation indicators, the 12 most predictive variables were then selected using XGBoost, including the participants' relationship status, sexual knowledge, sexual attitude, and previous sexual experience. Through MLR, RSB was found to be significantly associated with less sexual knowledge, more liberal sexual attitudes, single relationship status, and increased sexual experience. CONCLUSIONS: RSB is prevalent among college students. The XGBoost model is an effective approach to predict RSB and identify corresponding risk factors. This study presented an opportunity to promote sexual and reproductive health through ML models, which can help targeted interventions aimed at different subgroups and the precise surveillance and prevention of RSB among college students through risk probability prediction.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".