Skip to content

Identifying the origin of fish from the growth-ring diameter of scales using Neural Networks

Notifications You must be signed in to change notification settings

danieljordan2/Classifying-fish-using-NN

Repository files navigation

Classifying-fish-using-NN

This is a simple case using real data and was solved using a single neuron perceptron.

“Environmental authorities concerned with the depletion of salmon stocks decided to regulates the catches. To do this, it is necessary to identify whether a fish is of Alaskan or Canadian origin. Fifty fish from each place of origin were caught, and the growth-ring diameter of scale was measured for the time they lived in freshwater and the subsequent time they lived in saltwater. The aim is to identify the origin of a fish from its growth-ring diameter in freshater and saltwater.”

image

Fig 1. The growth-ring diameter of salmon in freshwater and saltwater for Canadian (red) and Alaskan (blue). Source: Samarasinghe S. [1]. Own elaboration.

Objective

Identifying the origin of fish from the growth-ring diameter of scales using different Neural Network algorithms.

Results

image

Fig 2. Samarasinghe's classifier (yellow, 96% acc), my classifier using Hebb rule (green, 96% acc), and my classifier using the delta rule (blue, 92% acc) superimposed on the data. Source: Samarasinghe S. [1]. Own elaboration.

Case study and data source

[1] Samarasinghe, S. Neural Networks for Applied Sciences and Engineering. From Fundamentals to Complex Pattern Recognition, Auerbach Publications, pp.35-40, 2007.

About

Identifying the origin of fish from the growth-ring diameter of scales using Neural Networks

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages