Forecasting Macroeconomic Parameters with Deep Learning Neural Networks

Topics Discussed Data Variables Meaning of variables in the data set CPI: Consumer price index. GDP: Gross domestic product UR: unemploymeent rate. IR Policy Rate - Interest rate (Possibly inflation subtracted from  nominal interest rate) LR10: Possibly 10 year loan rate. LR10 - IR: IR subtracted from LR10. Exrate Euro for 1 USD - exchange rate b/w euro and dollar. Wheather or not data has been preprocessed in any way - about to find out. Data analysis and preprocessing Should check if data is stationary or not. Identify any trend and seasonality. Test significance of each variable using a benchmark architecture (simple MLP). Explore different preprocessing techniques and if they improve performance on benchmark architecture. Longer term question - do we need more data?  Report Possible structure of Problem Analysis? Data - what kind and where froerform m? Stationarity, Seasonality, Trend Neural network architecture Tasks Completed Rea...

Topics discussed Literature Review Basic Structure (for now) Macroeconomics Importance of macroeconomic forecasting. Analysis of the data set and different parameters. Time Series Analysis of standard forecasting approaches. Discussion on seasonality, stationarity and other characteristics of time series data. Deep Learning Discussion of the emerging field of deep learning - advantages and disadvantages. Analysis of deep learning approaches. Macroeconomic Forecasting using Deep Learning Connecting the above topics and analysis on the work done in the field so far. Missing Data Dealing with missing values of parameters in the data should be done with caution. Possible solutions include leaving out data points with missing values or imputation techniques to estimate missing values. Leaving out data points may cause disruption in the continuity ofe the time series. Reading Resources Using R for Time Series Analysis  - Classical methods for time series analysis a...

Work done for past week Created project report outline in Latex. Created private GitHub repository for project. Started exploring Keras. Created reading list . The list includes all papers, books and past projects that I have read so far and will be continuously updated. Tasks for next week Read more on deep learning. Find research papers on: financial forecasting, time series prediction, deep learning, macroeconomic forecasting. Gain experience with Keras. Figure out correct formatting for project report (fonts, margins etc.) and modify document accordingly. Reading List Report Outline

Report Structure Introduction: Introduce topic in simpler terms. State goal, objectives and evaluation criteria. List chapters and explain structure of report. Only here it is acceptable to write about yourself(e.g. motivation to do this project). Goals should be to find out weather or not some hypothesis is true and it shouldn't be "make a better algorithm".  Literature Review Review what has been done so far in field of research. It is useful to make a table of all read papers, which summarises their most important points for the project. Problem Analysis Explain problem and argue why you have chosen solution (that you've chosen) based on the literature review. Software requirements go here. Design and Implementation Discuss the design and implementation of your solution. Results and Evaluation State results. Discuss what they mean and conclude which is the best one. Conclusion Draw conclusions. Can include speculations here (personal opinion). Suggest fu...

Outcomes Discussed some reading resources. Established which data set is going to be used. Discussed means of communication - through blog. Discussed some past projects and context of the current project. Scheduled next meeting to discuss project structure. Tasks Create Blog Read past projects connected to macroeconomic forecasting and/or deep learning  Read paper from project description Familiarise myself with the data Start using Keras