The most popular and comprehensive Open Source ECM platform
Let us know more
Artificial Intelligence: The Building Blocks of Neural Nets
Neural networks are algorithms that are able to make a decision after been trained by processing many similar examples. For example, an algorithm might be trained for image recognition of different types of objects by processing many images that label the types of objects shown in the image. The algorithm looks for patterns and differences based on the object labels.
Implementations of neural network algorithms with specialized hardware are referred to as ‘deep learning‘. The area of deep learning is being widely researched and there are thousands of different neural network algorithms being studied.
The different types of neural network algorithms can be generally classified into three groups:
- Multilayer Perceptrons (MLPs)
Typically used for data classification and natural language processing tasks like speech recognition and machine translation. - Convolutional Neural Networks (CNNs)
Typically used in computer vision for classification, segmentation and object detection of images - Recurrent Neural Networks (RNNs)
Typically used for analyzing time series data for processing sequences that occur in text, speech and videos. Used for classification and prediction.
Frank Emmert-Streib, Associate Professor at Finland’s Tampere University of Technology, wrote that “given the flexibility of network architectures allowing a “Lego-like” construction of new models, an unlimited number of neural network models can be constructed by utilizing elements of the core architectural building blocks discussed in this review. Hence, a basic understanding of these elements is key to be equipped for future developments in AI.”
0 comments on “Artificial Intelligence: The Building Blocks of Neural Nets”
1 Pings/Trackbacks for "Artificial Intelligence: The Building Blocks of Neural Nets"
-
[…] Read more at formtek.com […]