Carlos obtained his master’s degree in physics in 2017 at Universidad Veracruzana. He is currently working with LANIA helping companies make better use of the data they collect by telemetry. During his graduate studies, he developed an interest on econophysics, and he is interested in how artificial neural networks might play a crucial role in modeling and predicting economic systems. In his spare time, he enjoys cycling, video games and making animation videos.
Project: Electricity Price Forecasting
Goal of the project:
Find the variables and indicators relevant to the pricing of electricity in each state of the US, and from that data construct a predictor able to forecast prices.
Summary of work:
There is a lot of information involved in pricing electricity. Half of the work done was identifying the relevant data and working with the multiple governmental databases. The main variables found to be related to pricing electricity are the grid from which the electricity is taken, the heating and cooling degree days (which are a measure of electricity consumed because of cold and hot days), the GDP of the state and the price of thermal coal and natural gas.
Results and future work:
From this data a predictor was made, with a standard deviation of 0.505262 cents per kilowatt hour and a 65.7% success rate in predicting the sign of the return. This result may be improved by considering more data about the grid or related variables.