The AI behind Predict generates eye-tracking results with over 95% accuracy.
The accuracy of Predict AI
The Predict Attention AI has trained over 30 different machine learning models, each model running for weeks to months. The models were built on one of the world’s largest single databases of high-quality eye-tracking data. The winning model predicts eye-tracking results with over 95% accuracy. Read more about how we built Predict in the Technical Paper.
The scientific method behind Predict
Predict is an AI tool that is based on our world-leading consumer neuroscience database. With data based on real consumer responses, we are producing highly accurate predictive models of attention, and cognition.
With a database based on well over 20.000 participants, Neurons holds one of the largest single databases of high-quality eye-tracking data. The database involves properly labeled, quality eye-tracking recordings on consumer-related items (e.g., print ads, commercials, web pages, e-commerce, products, packaging, retail, and apps). This is a critical element in training machine learning models.
A variety of machine learning models were trained and compared to produce the best possible model prediction. For each model training, one portion of the data was randomly selected for training and a second portion for a validation test. The model was created by leveraging the latest versions of Google’s TensorFlow framework in order to create a fully convolutional autoencoder. The architecture employed structural elements from ResNet50 and MSINet as well as LSTM modules in order to create a balanced probability map.