Preprocessing Data And Fast Fourier Transform by Thom Hartle
Preprocessing data is a popular buzzterm today, but what is it and why do we do it? Well, here's an explanation of data preprocessing and a tutorial on using fast Fourier transforms (Fft). Ffts are used to measure the power and frequency of the cycles within data. As Stocks & Commodities Editor Thom Hartle explains, knowing the cyclical nature of your data can be very helpful for selecting parameters for your favorite indicators.
Every day, technicians process price data. If you use a moving average of the price, then you are
smoothing or reducing unwanted noise in the data to identify the trend. If you are using an oscillator such
as the relative strength index (RSI), then you are removing the trend in the price to aid in identifying
possible price extremes. Simply, processing data is mathematically transforming the data from one form
into another with the goal of amplifying the pertinent information for traders. Most often, the processed
data is then presented on a chart for generating trading signals in markets. In short, the price data has
been preprocessed before the trader analyzes the data visually.
PREPROCESSING AND ANALYSIS
Preprocessing therefore implies that the data can be more accurately analyzed after it's been altered to
some extent. For most people, it is simply easier to look at an indicator and form a conclusion than to
look at the price data itself. Market data appears to have a random or noisy appearance, and processing
the data into another form can help reduce the noise. Occasionally, a technical indicator or tool has been