Machine learning-based ABA treatment recommendation and personalization for autism spectrum disorder: an exploratory study
Machine-learning software can draft ABA goals that match clinician plans and lead to mastery, giving you a speedy, evidence-based starting point.
01Research in Context
What this study did
Kohli and team built a computer program that writes ABA goals. The program learns from past treatment records. It then suggests new targets for each child.
Clinicians compared the computer goals to their own plans. They also tracked how many suggested goals kids actually mastered.
What they found
The machine's goals matched clinician plans 81-84% of the time. Kids went on to master most of the computer-picked targets.
The tool did not replace the BCBA. Instead, it gave a fast first draft that experts could tweak.
How this fits with other research
Kausch et al. (2026) also use machine learning on ABA data. They teach computers to read ATD graphs. Both papers show ML can handle routine visual tasks, freeing you for clinical decisions.
Gitimoghaddam et al. (2022) scoured 770 ABA studies. They found almost no controlled trials and zero quality-of-life data. Kohli’s work fills that gap by adding automated, data-driven goal selection.
Cerasuolo et al. (2022) warn that no single child trait guarantees success. The new ML tool sidesteps that trap. It weighs dozens of past variables at once instead of betting on one predictor.
Why it matters
You can try an ML helper today. Export redacted client data, let the model draft goals, then review and adjust. You keep clinical control while cutting prep time. More kids get personalized plans, and you get evidence the suggestions work.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Run a pilot: feed last quarter’s mastered and non-mastered goals into an open-source ML tool, generate draft targets for one client, and compare them to your current plan.
02At a glance
03Original abstract
Autism spectrum is a brain development condition that impairs an individual’s capacity to communicate socially and manifests through strict routines and obsessive–compulsive behavior. Applied behavior analysis (ABA) is the gold-standard treatment for autism spectrum disorder (ASD). However, as the number of ASD cases increases, there is a substantial shortage of licensed ABA practitioners, limiting the timely formulation, revision, and implementation of treatment plans and goals. Additionally, the subjectivity of the clinician and a lack of data-driven decision-making affect treatment quality. We address these obstacles by applying two machine learning algorithms to recommend and personalize ABA treatment goals for 29 study participants with ASD. The patient similarity and collaborative filtering methods predicted ABA treatment with an average accuracy of 81–84%, with a normalized discounted cumulative gain of 79–81% (NDCG) compared to clinician-prepared ABA treatment recommendations. Additionally, we assess the two models’ treatment efficacy (TE) by measuring the percentage of recommended treatment goals mastered by the study participants. The proposed treatment recommendation and personalization strategy are generalizable to other intervention methods in addition to ABA and for other brain disorders. This study was registered as a clinical trial on November 5, 2020 with trial registration number CTRI/2020/11/028933.
Brain Informatics, 2022 · doi:10.1186/s40708-022-00164-6