Support vector machine algorithm for separable data using elementary geometry

A fast iterative support vector machine algorithm for linearly separable dataset. Makes use of elementary geometry; no quadratic programming has been employed.

• provides weight and bias that defines the optimum-hyperplane and also gives the support vectors.
• applicable only for linearly separable problem. i.e., not suitable for soft-marging classification
• if the data is not separable, the algorithm quickly figure out this and stops further iterations.