Hi Professor, this is a very fascinating topic to me as to applying economic models in historical studies. Since the data is about trade travels, which dated back to four thousand years ago, I have noticed that although we have about 12000 available texts, only 391 directed travels between two cities are available to be analyzed. As a result, the average number of trade travel for any trading pairs is only 3.26 and the data is very dispersed. Compared to the data collected nowadays in digital age, the data size of your model is very small and at first glance I suspect if it can deliver any informative predictions through your model. I am curious about how much information you think your small data size can provide or how accurate the results are from your model? If the data size available is much larger, do you believe the predictive results of lost city locations and ancient city populations will deviate substantially from your current predictions? Thank you so much!
Hi Professor, this is a very fascinating topic to me as to applying economic models in historical studies. Since the data is about trade travels, which dated back to four thousand years ago, I have noticed that although we have about 12000 available texts, only 391 directed travels between two cities are available to be analyzed. As a result, the average number of trade travel for any trading pairs is only 3.26 and the data is very dispersed. Compared to the data collected nowadays in digital age, the data size of your model is very small and at first glance I suspect if it can deliver any informative predictions through your model. I am curious about how much information you think your small data size can provide or how accurate the results are from your model? If the data size available is much larger, do you believe the predictive results of lost city locations and ancient city populations will deviate substantially from your current predictions? Thank you so much!