Neural Network Applications in General Science

Neural networks provide significant benefits in general science research. They are actively being used for such applications as satellite imagery, meteorology, detecting terrorist activities, and more.

Below we have listed some of the common applications of neural networks in general science research. If you are currently using neural networks in your general science application, we would love to hear about it.

NeuroDimension has also used its leading edge neural network technology to develop numerous science applications with a variety of companies. If you need neural network consulting for your science application, please contact NeuroDimension.

Sample Applications

Locate common characteristics in large amounts of data (divide research populations).
Better forecast results based on existing data (emissions, changes to device settings).
Predict the progression of scientific data over time (erosion, probability of mechanical failure).
Identify characteristics in images or video feeds (traffic flow, target detection, machine status).
Group scientific data based on key characteristics (robotic movements, forestry type).

Locate common characteristics in large amounts of data.

Locating common characteristics in large amounts of data is a type of classification problem. Neural networks can be used to solve classification problems, typically through Multi-Layer Perceptron (MLP) and Support Vector Machines (SVM) type networks.

Examples of classification in general science include classifying research populations into groups. For example, data from studies can potentially help create advance warning systems in the future for land slides or even tsunamis, earthquakes and other natural disasters.

Sample Study: Classification of satellite imagery for detecting land slides

Using NeuroSolutions and satellite imagery this customer was able to successfully classify 85% detection for landslide and 73% for non-landslide. This study also concluded that it is feasible to gain a synergy of information on high resolution images, digital terrain models, existing roads and drainage systems and automate the information for landslide identification. Landslide Features Interpreted By Neural Network Method Using A High-Resolution Satellite Image and Digital Topographic Data - K. T. Chang and J. K. Liu Locate this paper on Google Scholar!

Sample Project: Iris Classification.

Using NeuroSolutions, this project made famous by Fisher, who used it to illustrate principles of discriminate analysis. It contains 4 input variables, the petal width and length and the Sepal width and length. The task is to use these inputs to classify the Iris into one of 3 Iris species. Download the dataset: Iris.zip, 508KB Use with NeuroSolutions: Free Evaluation

Customer Interview: Paul Roebber

Mr. Roebber has been using NeuroSolutions along with the Custom Solution Wizard and Source Code Library for the past 8-years. Using NeuroSolutions, Mr. Roebber developed a snow density system that is currently employed by the National Weather Service to assist in the prediction of snow fall depths. For the complete Customer Interview visit NeuroSolutions.com!

Our NeuroSolutions product is an excellent resource for classification applications. For an interactive example of classification in NeuroSolutions for Excel, download the free evaluation version and view the demo called “Testing Classifiers” in the Help menu.

Better forecast results based on existing data.

Forecasting the relationship between multiple factors in general science data is a type of function approximation problem. Neural networks can be used to solve function approximation problems, typically through Multi-Layer Perceptron (MLP), Radial Basis Function (RBF) and CANFIS (Co-Active Neuro-Fuzzy Inference System) type networks.

Examples of function approximation in general science include predicting changes to device settings and emissions. For example, data from studies can potentially help predict meteorological applications such as minimum temperature, fog, rainfall estimation forecasting and many more.

Sample Study: Meteorological Applications in Neural Networks.

This sample study highlights twelve neural network based meteorological articles ranging from lightning to rainfall forecasting using neural networks. Some of the articles reveal a very high success rate on forecasting certain events like “Cloud Recognition with Ground Sensors”, “Lightning Forecasting” and many more. State of the Art of Neural Networks in Meteorology - Bjarne Hansen Locate this paper on Google Scholar!

Sample Project: Determining Fuel Economy.

Using NeuroSolutions, this project uses a dataset based on vehicle information such as cylinders, horsepower, weight and so on to determine fuel economy (MPG - Miles per Gallon) for each vehicle. There are a total of 392 samples in this data set with 9 input variables and 1 output. Download the dataset: MPG.zip, 507KB Use with NeuroSolutions: Free Evaluation

Our NeuroSolutions product is an excellent resource for function approximation applications. For an interactive example of function approximation in NeuroSolutions, download the free evaluation version of the software and view the demo called “Multi-Layer Perceptron, Basic” in the Help menu.

Predict the progression of scientific data over time

Forecasting the relationship between multiple factors in general science data is a type of time-series prediction problem. Neural networks can be used to solve time-series problems, typically through Time-Lagged Recurrent (TLRN) type network.

Examples of time-series predictions in general science include forecasting flood predictions. For example, data from general science studies can model the river flow to provide real-time prediction of peak downstream.

Sample Study: Real-time River Flood Prediction

Using NeuroSolutions, the study area covers the Upper Derwent River, a tributary of the River Trent in the United Kingdom. Tests were completed using different lengths of input data to evaluate the effect of input data size in model outputs. According to the results of this research it can be said that for real-time forecasting of flow in gauged catchments the type of neural network is an important factor and dynamic architectures, especially general recurrent networks, show a superior ability even for longer prediction horizons. Evaluation of the application of neural networks on real-time river flood prediction - M.T. Dastorani Locate this paper on Google Scholar!

Our NeuroSolutions product is an excellent resource for time-series prediction applications. For an interactive example of time-series prediction in NeuroSolutions, download the free evaluation version of the software and view the demo called “Time Lagged Recurrent Network” in the Help menu.

Identify characteristics in images or video feeds

Identifying characters in images or video feeds in general science is a type of image processing problem. Neural networks can be used to solve image processing problems, typically through Principal Component Analysis (PCA) type network.

Examples of image processing in general science include identifying agents in images. For example, image data from general science studies can model the detection of lethal agents using Terahertz images.

Sample Study: Detecting Lethal Agents from Images

Using NeuroSolutions, the study used non-invasive means to detect and characterize lethal agents using spatial imaging of their characteristic transmission or reflection wavelength spectrum in the Terahertz (THz) electro-magnetic range. The application was put to use in two forms in this study: Lethal Agent detection in an Envelope and Suicide Bomber. Neural Network Analysis of Interferometric Terahertz Images For Detection of Lethal Agents - F. Oliveira and R. Barat, Otto York Department of Chemical Engineering and B. Schulkin, F. Huang, J. Federici, and D. Gary, Department of Physics Locate this paper on Google Scholar!

Sample Study: Seabed Recognition Using Neural Networks

Using NeuroSolutions, this study covers data exploration of sonar image segments has been carried out using SOM. The developed seabed recognition system consists of a tool for feature extraction from sonar images and two neural network classifiers, the labeled SOM feature map and SOM_MLP classifier. The system identifies the seabed materials like clay/mud, sand, eel grass and gravel from images using five selected features of the image segments; median, 3rd quartile, energy, entropy and momentum. Seabed Recognition Using Neural Networks - Vladan Babovic Locate this paper on NeuroSolutions.com! Additional Sample Studies: Path Recognition for Military Robots

Sample Project: Texture Classification: Using Neural Networks to Differentiate a Leopard from its Background

Using NeuroSolutions for Matlab, this problem is to distinguish between the leopard and the background, in which it is sitting, in the image shown below. This problem falls under the category called texture classification, which falls under the broader category of pattern classification. Click here for complete details. Download the dataset: tex.zip, 161KB Use with NeuroSolutions: Free Evaluation

Customer Interview: Mats Rosengren

Mr. Rosengren has been using NeuroSolutions and NeuroSolutions for Excel for the past 6 years and has also attended one of our Neural Network Courses in Orlando, Florida. Using remote sensing satellite imagery, multi-spectral sensor data is used for applications to map forest types or changes in a forest or to extract statistical information from the images together with available maps. In the case of mapping of forest types and changes, the information is essential for forestry companies in order to keep their databases up-to-date for planning and forest management. In addition, the imagery has been used to do wall-to-wall mapping of land use and vegetation from space as well as measuring the depth in the sea water, mapping along the coastlines and measuring the vegetation at the bottom of the sea. For the complete Customer Interview visit NeuroSolutions.com!

Our NeuroSolutions product is an excellent resource for image processing applications. For an interactive example of image processing in NeuroSolutions, download the free evaluation version of the software and view the demo called “Linear Associator” in the Help menu.

Group scientific data based on key characteristics

Grouping of general science data based on key characteristics is a type of clustering problem. Neural networks can be used to solve clustering problems, typically through Self-Organizing Map (SOM) type network.

Examples of clustering in general science include the detection of key characteristics in robotic movement and forestry separation. For example, data from studies concerning machine learning can be applied to help robots improve their operational capabilities.

Sample Study: Neural Networks in Mobile Robot Motion

Using NeuroSolutions, this study deals with a path planning and intelligent control of a autonomous robot which should move safely in partially structured environment. The study is based on two neural networks: 1) Determine “free” space using ultrasound range finder data and 2) “Finds” a safe direction for the next robot section of the path in the workspace while avoiding the nearest obstacle. Neural Networks in Mobile Robot Motion - Danica Janglová Locate this paper on Google Scholar!

Our NeuroSolutions product is an excellent resource for clustering applications. For an interactive example of clustering in NeuroSolutions, download the free evaluation version of the software and view the demo called “Unsupervised Learning” in the Help menu.