Reference Use Cases
Analytics Zoo provides a collection of reference user applications and demos, which can be modified or even used off-the-shelf in real world applications. Some are listed below. See all in analytics-zoo/apps.
-
Anomaly Detection demostrates using LSTM network to detect anomalies in time series data.
-
Fraud Detection demostrates using feed-forward neural network to detect frauds in credit card transactions data.
-
Image Augmentation demostrates how to do image augmentation for vision projects.
-
Object Detection demonstrates how to use Analytics Zoo Object Detection API (and pretrained SSD model) on videos.
-
Recommendation demonstrates how to use Analytics Zoo Recommendation APIs (i.e. Neural Collaborative Filtering, Wide and Deep) to do recommendation on data with explicit feedback.
-
Sentiment Analysis demostrates how to do sentiment analysis using neural network models (e.g. CNN, LSTM, GRU, Bi-LSTM).
-
Variational AutoEncoder demostrates how to use variational autoencoder to generate faces and digital numbers.