-
Book Overview & Buying
-
Table Of Contents
Hands-On Artificial Intelligence for IoT - Second Edition
By :
In this chapter, we covered what time-series data is and how it differs from other tabular data. We explored SparkML and used its built-in DecisionTreeRegressor to predict CO levels using the air quality dataset from the UCI Machine Learning Repository.
Next, we recapped RNNs and LSTMs, discussing why they are well suited for time-series forecasting. We built an LSTM model using Keras and trained it on the same air quality data.
We then shifted to pretrained models, fine-tuning Prophet on the dataset for more efficient predictions. Other pretrained models, such as Chronos and Hugging Face Transformers, were briefly discussed in terms of accuracy and edge deployment.
Throughout the chapter, we focused on the practical aspects of edge deployment, data cleaning, and model optimization to ensure efficient real-time forecasts on constrained devices.
In the next chapter, we will work with video and image data from IoT devices, exploring techniques for real-time visual...