Estimating the values of dollar index, Nasdaq index, gold and Bitcoin depending on each other by machine learning methods

Investment is an important resource for people. Individuals want to use their savings with financial instruments. There are many financial instruments available today, such as the dollar, gold, euro, silver, stock market and, recently popular Bitcoin and other cryptocurrencies. Machine learning is used in many fields. In the field of finance, machine learning is used to predict the future by making use of past data. In this study, the dollar index, which shows the strength of the dollar against major currencies, the Nasdaq 100 index, which hosts non-financial companies in the United States, the ounce value of gold, one of the oldest and most valuable investment instruments, and the value of Bitcoin, which has been increasing in popularity recently, have been used to help investors and make forward-looking predictions. The biennial, three-year and ten-year data were created with the data taken from Investing.com website, in the light of closing values for the years 2012-2022. Linear Regression, Lasso and Decision tree models are used in machine learning with Python programming language. Three currencies were made as independent variables, and the dependent variable currency was estimated. As a result of the study, it was observed that the application run with ten-year data set gave more successful results, and as a method, Lasso and Decision Tree models obtained successful results from linear regression. It has also been revealed that the application tested with real values that are not in the data set gives results close to the real values.