A Convolutional Neural Network (CNN or ConvNet) is a way to implement an artificial neural network. CNNs are used in numerous modern artificial intelligence technologies, especially in the machine processing of sequential data sets, but also in images. Basically, the structure of a classical neural CNN consists of one or more convolutional layers, followed by a pooling layer and a fully-connected layer. Basically, these units or layers can be repeated as often as desired. By learning and applying features (training) it is possible to recognize patterns and objects over the network. Higher layers use the data of the lower layers so that even more complex patterns can be recognized. Most use cases for 1D CNN models are those where sequential data sets are used. These are often sensor data measured at a defined time interval. For example, temperature, humidity or wind force are read in and analyzed.

The 1D-CNN was implemented as part of the lecture "Deep Learning" (DL) of the master's degree "Informatik" at Hochschule Kaiserslautern - University of Applied Sciences, Campus Zweibrücken.

https://www.hs-kl.de/

The first notebook discusses the use of 1D Convolutional Neural Networks (1D CNNs) to classify text in Keras. First, there is a brief introduction to this type of neural network and then shows the differences between a one-dimensional CNN and a two-dimensional CNN.

In the second notebook a 1D-CNN is deepened by a practical example. The example, which will be examined in more detail below as well as the corresponding data sets originate from a competition of the platform Kaggle. Kaggle is an online community of Data Scientists and allows users to find or publish data sets. The platform also allows users to explore or create models in a web-based data science environment, collaborate with other data scientists and engineers, and compete to solve data science challenges. By conducting machine learning competitions, a public data platform, a cloud-based workbench for data science, and a short form of CI education, Klagge offers its users a lot of interesting content in this area.

Please note that the training data sets required to train the model in the second notebook are not available in this repository. However, you can find and download the datasets under the following link: https://www.kaggle.com/c/LANL-Earthquake-Prediction/data

• 1.0: Veröffentlichung

Alea Ilona Sauer – GitHub Profil
Eric Gustav Werner – GitHub Profil