Project - Maintaining ecological balance using machine learning algorithms

TOPIC:
     MAINTAINING ECOLOGICAL BALANCE USING MACHINE LEARNING ALGORITHMS

MOTIVE:

Our environment has a lot of plant and animal species and each one is obviously dependent on one other for its survival. The animals living in a particular ecosystem suffer when there is an increase or a decrease in the number of species that must actually exist in the ecosystem. This is because if there is an increase in a particular kind of species in the ecosystem, there might not be sufficient food available for all of the animals to exist and at the same time if there is a decrease in a particular kind of species, these species might be the food for some other animal in the ecosystem. As a result, those animals will be left without food, thereby affecting the food chain. Thus, the number of species that exist in a particular system has to be maintained at a constant rate to maintain and preserve ecological balance.  So, our group thought that if there exists some automated system, using the concept of machine learning, which could actually monitor the current population in a particular ecosystem, it will be of great help in maintaining ecological balance.

ALGORITHM:

Machine learning algorithms can be used to easily solve real life situations, using computer algorithms.  Machine learning algorithms predict the output by establishing by a pattern in the given set of inputs. The inputs are nothing but training sets. The population in a particular ecosystem is first stored in the database.  The ecosystem should be constantly monitored and any increase or decrease of number of the species living in the ecosystem should be constantly updated in the database. This database actually acts as the input for the machine learning algorithm. The machine learning algorithm already would have established some patterns in the input when the database was first created. That is, suppose if we take pond ecosystem for example, let us assume that it has A,B,C and D species living in it. Classification rules can be used to classify the population of species A, species B, species C and species D so that the population count of all the species is correctly maintained and any imbalance in the ecosystem is correctly detected. The optimal number of species which must be available in the ecosystem is already stored in the database. Machine learning algorithm should be designed in such a way that whenever there is a rise in population of any particular species in the ecosystem, it can easily detect by comparing the input patterns. Decision trees are used to find whether the population is within limits or whether it has exceeded the limits. If it is within limits then the algorithm does nothing. But if some change is detected, then this information can be intimated to some governmental organization who can take the necessary actions to transfer those species to another ecosystem, so that the ecological balance in that ecosystem is maintained.