Contingency table.md

Contingency table

A contingency table shows all the classifications that have been done with the real value. In a contingency table, you can see exactly how things were wrongly classified as what.

Example

Here is an example table in a result when classifying languages.

Language	German	French	Dutch	Italian	English
German	4030.	1.	2.	4.	13.
French	3.	3385.	2.	11.	7.
Dutch	20.	2.	1230.	4.	24.
Italian	0.	4.	2.	10682.	4.
English	10.	20.	13.	43.	16559.

So what you can read from this table is that 4030 German words were classified as German 4030 times but also 3 French, 20 Dutch and 10 Italian words were classified as German.

1 German word was also classified as French. 2 German words were also classified as Dutch. 4 German words were also classified as Italian. 13 German words were also classified as English.

So that is how you read these tables.

Here is example code to create the evaluating classification models scores.

def perf_measure(y_actual, y_hat):
    TP = 0
    FP = 0
    TN = 0
    FN = 0

    for i in range(len(y_hat)): 
        if y_actual[i]==y_hat[i]==1:
           TP += 1
        if y_hat[i]==1 and y_actual[i]!=y_hat[i]:
           FP += 1
        if y_actual[i]==y_hat[i]==0:
           TN += 1
        if y_hat[i]==0 and y_actual[i]!=y_hat[i]:
           FN += 1

    return TP, FP, TN, FN

def precision(y_actual, y_hat):
	TP, FP, TN, FN = perf_measure(y_actual, y_hat)
    return TP / (TP + FP)


def recall(contingency_table, label):
	TP, FP, TN, FN = perf_measure(y_actual, y_hat)
    return TP / (TP + FN)

def F_measure(y_actual, y_hat, Beta=1):
    P = precision(y_actual, y_hat)
    R = recall(y_actual, y_hat)
    return ((Beta**2 + 1) * P*R) / (Beta**2 * P + R)