CHAPTER - 6

6.1 EXPERIMENTAL RESULTS OF VOICE DEFLECTIONS

We have used several measures to evaluate the predictive results. These measures are Confusion Matrix, Accuracy, Precision, Recall or Sensitivity, Specificity, F1-score.

For machine learning processes, a classification model's or algorithm's performance is represented using the confusion matrix, which can be either a chart or a table. The confusion matrix is also called as Error matrix. It is a table that is often used to describe the performance of a classification method on a set of test data for which actual value are known. Each column of the matrix represents the instances in a predicted class.

Confusion matrix is the useful tool for determining which tasks a machine learning system executes correctly and incorrectly. It includes both the systems predicted and true values when constructing a confusion matrix. Each predicted values has a related row and each true values has a related column.

A confusion matrix presents a table layout of the different outcomes of the prediction and results of a classification problem and helps visualize its outcomes. It plots a table of all the predicted and actual values of a classifier. The Confusion matrix for the voice data analysis using the XGBoost Algorithm is shown in the below figure. The figure describes that the predicted healthy persons are 22 and the PD persons are 6 in the dataset.

In the field of machine learning and specifically the problem of statistical classification, a confusion matrix, also known as error matrix, is a specific table layout that allows visualization of the performance of an algorithm, typically a supervised learning one; in unsupervised learning it is usually called a matching matrix.

It is a special kind of contingency table, with two dimensions ("actual" and "predicted"), and identical sets of "classes" in both dimensions (each combination of dimension and class is a variable in the contingency table). The different classification metrics like accuracy, sensitivity, specificity, precision, recall and recall of the proposed algorithm are measured and is shown in the below figure. The accuracy is measuredusing the different machine leaning algorithms with the speech data as input.

The testing of the data is done by using the data from the dataset and give that data as input. Fromthe data except the status column has given to the input which includes 22 parameters. The numpy array has given to the scaler transform which will minimize and reshape all the values in the range (1,-1). Afterscaling the input the model will predict the disease. If the output is ‘0’ it says that the person does nothave the PD, else it says the person is having the PD.

6.2     EXPERIMENTAL RESULTS OF SPIRAL DESIGNS

The different classification metrics like accuracy, sensitivity, specificity, precision, recall and recall of the proposed algorithm are measured. These metrics are measured using the different machine leaning algorithms with the spiral drawings data as input which will give the predictions of the given data. Here below we can see the Model accuracy and Model loss graphs of the wave and spiral drawings.

The comparison of the different machine learning algorithms with the present algorithm to detect the PD based on the handwritten drawings is shown in the below Table-II. Table II contains the different parameters which are used to check the present used machine learning algorithm is better or not. The comparisons made from the previously used algorithms with the proposed algorithms is shown in the below table II. From the figure and the table II, the proposed model is having the better accuracy which is 86.67% and it is considerable compared to the other previously used algorithms.