Task Summary: Validation for LBP and SVM methods. Created a manual calculation simulation that proves my system's workflow, from image input to classification decision. The system uses the following methods: Feature Extraction: Local Binary Pattern (LBP). Classification: Pre-trained Support Vector Machine (SVM). IMPORTANT: You will NOT be asked to retrain the model or recode it. Your task is to analyze the mathematics behind the model's predictions for just one data sample so that it can be written in the report. Data & Files I Will Provide: -Original Video File (.mp4): I will provide the video. -Validation Matrix: For manual calculations in Excel, you can simply sample a small area (5x5) of the face to demonstrate the LBP (Thresholding & Binarization) formula. There is no need to calculate a full image pixel-by-pixel, as that would be too much. -SVM Validation: Use the pre-made feature values (histogram) (from my CSV file or your simulation results) to input into the manual SVM formula. -Trained SVM Model: Model file. Scope of Work: You must present the step-by-step manual calculations in an Excel file for the following stages: Stage 1: LBP Feature Extraction (Manual) -Pixel Operations: Demonstrate the calculation of the median minus the neighboring values (Thresholding) in the matrix. -Binarization: Demonstrate the formation of binary bits (clockwise). -Decimalization: Convert binary to decimal to obtain the LBP values. -Histogram Formation: Demonstrate how the LBP values are calculated by frequency to form a Feature Vector (e.g., 1 row of an array containing 59 or 256 numbers). -Normalization: Demonstrate the division of the histogram values by the total number of pixels. Stage 2: SVM Classification (Manual) -Input: Use the Feature Vector generated from Stage 1. -Decision Function: Show the mathematical calculation: f(x) = (w . x) + b Where: w = Weights from my SVM model. x = Feature Vector from Stage 1. b = Bias (Intercept). -Dot Product: Show the multiplication of each feature element with its respective weight in Excel. Class Determination: Show the final logic. Example: If Result > 0 -> "Real", If Result < 0 -> "Fake". Expected Output (Deliverables): Excel File (.xlsx) - "The Calculator" -All cells must use ACTIVE FORMULAS (not dead numbers/manual typing). -Separate sheets for "LBP Process" and "SVM Process". -Easy to track: If I change the input pixel values, the final classification results in Excel should automatically change according to the formula. Document File (.docx) - "The Report" -A brief narrative explanation of the calculation steps performed in Excel. -Screenshots of key parts of Excel. -The format must be neat for adaptation. Acceptance Criteria: -The final results of the manual calculations (feature counts & SVM Decision Function values) must EXACTLY MATCH the output of my Python/CSV program. A decimal difference tolerance is allowed (0.0001). -Explain the kernel formula