This repository covers the work on time series forecasting that I have doen in the summers of 2022.
Using Times series and forecasting to predict opening price of a stock click
Performed time series analysis and forecasting of opening price of Tata Steel
Performed smoothing out of the irregular roughness to better see patterns, trends, and smooth out the seasonality to better identify the trend.
Moving averages are a simple and effective way to smooth time series data. They work by calculating the average of a fixed number of data points over a rolling window. This helps to remove high-frequency noise and highlight the underlying trend of the data.
There are two main reasons why moving averages are best for smoothing time series:
Simplicity: Moving averages are very easy to calculate and understand. This makes them a good choice for applications where interpretability is important.
Effectiveness: Moving averages are very effective at removing noise and highlighting trends in time series data. This is because they average over a fixed number of data points, which helps to cancel out random fluctuations.
Trimming time series data is important for noise reduction, improved analysis, data compression, focusing on relevant periods, handling missing values, enhancing visual representation, meeting statistical assumptions, and improving computational efficiency. click
The Holt-Winters model is frequently used for anomaly detection due to its effectiveness in handling time series data with trend and seasonality. Its ability to capture underlying patterns and identify deviations from those patterns makes it well-suited for detecting anomalies. click