search

FRI-Energy-Analytics / 2020_summer_fellows

A repository for FRI Energy Analytics fellows
MIT License
5 stars 6 forks source link

Solow Model #64

Open SsanyaS14 opened 4 years ago

SsanyaS14 commented 4 years ago

Where is the Model Used Economics

How does the Model work Solow Model looks at long-term economic growth. It factors in labor, capital accumulation, population growth, an increase in productivity, and technological progress. It is a non-linear system that can take into account the production function. You input production functions to understand the rate of growth.

Used in Oil We have multivariable data that needs to be distributed in a way to allow the growth or see the optimization of oil production. We can use all factors or try only a few to see how we can grow a certain optimization of oil.

samirabr12 commented 4 years ago

This model is interesting and one that I am excited to work with as the sprints go on. I know there is still more research to be done in order to completely understand this model, but if possible you can add the interpretation of each graph and what this tells us about oil production, in 1-2 sentences. Also, are you focusing on Spearman's Rank correlation, Pearson correlation, or both? I think you found the values for Spearman's Rank correlation coefficient but mentioned Pearson in the formula explanation, so I wasn't exactly sure about what was being used. But overall, we should definitely explore this model more starting next sprint because it has a lot of potential and can help us understand how different features contribute to the rate of growth for oil production.

samirabr12 commented 4 years ago

I think this model is one of the top models we have analyzed because it incorporates economic concepts in order to compare different features in our dataset and how they relate to cumulative production. For future work, I suggest trying out different correlation coefficients if you haven't officially decided to use Spearman's coefficient for this model. In addition, maybe incorporate the cost and time aspect of the data, which you mentioned earlier today. I'm not sure if the error can be measured with this model, but if it can be, then try to show an error measure or some way to show how accurate the model is. Also if possible, could you explain in the notebook your thinking behind replacing parts of the formula with certain features in the dataset, just so we can get a better understanding of the model? But overall, this model has a lot of potential and will be very helpful with feature selection and analyzing the data.

Shown Corr - A Shown Error - D Potential - A Interconnectedness - A Application - A