Study on classification algorithm of similar microalgae based on principal component features
Marine microalgae is a kind of photosynthetic autotrophic organism.With the increasing severity of algal bloom,algal blooms and red tides are formed gradually.Therefore,how to quickly and accurately i-dentify beneficial and harmful algae is an urgent task.In order to solve this problem that many algae are in-distinguishable from each other,the paper proposes several processes.Firstly,the input image is standard-ized and the first principal component load is calculated.Secondly,a binary classification model based on logistic regression is designed,and two kinds of labels are trained.Finally,the cost function and sigmoid function are constructed,and the accuracy of the gradient descent is 92.86％.Compared with the widely used binary algal classification algorithm SVM classifier,the result shows that the classification accuracy of similar algal is improved by 1.86％.
万方数据
维普
