Information Technology & Services - , ,
Most of the techniques in Statistics were developed previous to the Internet era: small datasets and very limited computational power. Although the computational power has been greatly improved in the recent years, it has not accompanied the growth of the size of data. Although much research is in progress, many techniques fail when facing current problems: they do not scale. Whether it is because of The Curse of Dimensionality or the lack of computing resources, many times the conclusions arrived from these models are meaningless. They are based on the know-how of the system, not on the model. While there are many problems than can be solved, there are many that can't be modeled properly at this moment. This is our challenge: - Select the appropriate techniques to your problem, adapt them and extend them if necessary. - Implement them. - Optimize them to be FAST. - Analyze the results based on the foundations provided by the model. - Provide the limitations of the model. - Solve the problem. We will analyze your problem and provide an outline of the solution at no cost. If we can't provide a reasonable optimal solution, we will try to provide a direction to find someone that can solve it. You can also read us at our blog: stathacking.com.