
-   Linear Algebra
-   Statistics
    -   Collect, Organize, Analyze, interpret & Represent
    -   Descriptive
        -   Measure of central tendency (MoCT)
            -   Mean (μ or x̄)
            -   Median (M)
            -   Mode (Mo)
        -   Measure of Dispersion (MoD)
            -   Range (R)
            -   Standard Deviation (s or σ)
            -   Variance (s² or σ²)
    -   Inferential
        -   Z-test
        -   T-test
        -   F-test
-   Probability
    -   Conditional Probability
        -   <img src="https://ik.imagekit.io/manav11goyal11/blog/images/machine-learning/tr:w-200/6.jpg" alt="Image Not Found" title="Conditional Probability" height="250px" />
        -   Find probability of event A given that event B has already occurred
        -   Baye's Theorem
            -   <img src="https://ik.imagekit.io/manav11goyal11/blog/images/machine-learning/tr:w-300/7.jpg" alt="Image Not Found" title="Baye's Theorem" height="250px" />
            -   A is Hypothesis and B is Data/Evidence
            -   Naive Bayes Classifier
                -   Bernoulli
                    -   Features are of Binary nature
                    -   Bernoulli distribution
                    -   P(success) = p & P(faliure) = q = 1 - p
                    -   Random Variable (X) => 1 (Success) & 0 (Faliure)
                    -   X has Bernoulli distribution
                    -   <img src="https://ik.imagekit.io/manav11goyal11/blog/images/machine-learning/tr:w-200/12.jpg" alt="Image Not Found" title="Bernoulli" height="250px" />
                -   Multinomial
                    -   Used for finding Discrete count, Number of occurrences, Frequency
                    -   Multinomial distribution
                    -   <img src="https://ik.imagekit.io/manav11goyal11/blog/images/machine-learning/tr:w-200/13.jpg" alt="Image Not Found" title="Multinomial" height="250px" />
                -   Gaussian (Normal)
                    -   Not to be used if data has discrete values, Use when features has continuous values
                    -   Gaussian distribution
                    -   <img src="https://ik.imagekit.io/manav11goyal11/blog/images/machine-learning/tr:w-200/14.jpg" alt="Image Not Found" title="Gaussian" height="250px" />
                    -   Code
                        ```PYTHON
                            from sklearn.model_selection import train_test_split
                            from sklearn.naive_bayes import GaussianNB
                            from sklearn import metrics
                            X = data.value
                            y = data.target
                            X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=1)
                            gnb = GaussianNB()
                            gnb.fit(X_train, y_train)
                            y_pred = gnb.predict(X_test)
                            print(metrics.accuracy_score(y_test, y_pred))
                        ```


Manav Goyal

Maths - Machine Learning