sjsufall2019-cmpe272 / rinnovation Goto Github PK

Support:

• Entering zip code
• Viewing line graph (x axis: year, y axis: ROI, lines: 1 for each renovation)

https://aws.amazon.com/blogs/compute/managing-secrets-for-amazon-ecs-applications-using-parameter-store-and-iam-roles-for-tasks/

Integrate @kaustubhkulkarni1509's React work into the Play service

The CSVs produced during the PDF table -> CSV conversion are not usable. They need to be massaged in order to be used by SKL.

• The use-cases are described in a good manner with what is going to be used as inputs and what are the expected outputs, but technologies are not yet finalised and thus the technology flow is not defined in this document.
• Thus there is no concrete architecture/flow we could comprehend from the use-cases mentioned above.

https://github.com/aws-samples/aws-cicd-docker-containers

Based on the definition of Cockroach DB in Wikipedia,

It is a project designed to store copies of data in multiple locations in order to deliver speed access.

Which is most suited for a web-application which has users across different parts of the world. The use of Cockroach DB in this machine learning use-case not a fit or needed.

• As I was reading through the description of your project, I was astonished by the number of technologies. I was amazed by the idea of implementation of technologies. But, as I read in detail about all these technologies, I got to know there is a lot of integration required to implement all these together. These technologies form a wide spectrum and are significant in their own aspects. But I think, in such a project, are all these technologies necessary? I thought that you have bloated too much about technologies.

• As an example, you can take about the service module, you are using Docker and AWS together. Instead, you could use Amazon EC2 and deploy your project in an auto-scaled and load-balanced environment. Also, in ETL you guys have given a lot of technologies. How will integrate all those? Each ETL technology has its own pros, cons and implementation methods. Hence, I think that in later phases of the project, you might get a lot of integration issues. You have not directly mentioned which ETL technologies you will use. You are a little shaky that you can try these and find out if it can work in your application. This will be time-consuming for your project.

• Hence, I think that your idea is really but, for implementation you have adopted a pretty wide approach, this might be time-consuming and might have a lot of integration issues.

• The customers might be having their own ideas of how to renovate their houses. The trained model might provide suggestions which are in trend, but not necessarily suit the taste of the customer. The customer might get value for the money only if they are completely satisfied.

• Hence, there should be a provision to consider customer's inputs of their likes and dislikes and suggest the design accordingly.

Using Akka Streams, to improve performance

The abstract of the project mentions that the machine learning model can predict the time and cost estimates of the renovation to be completed. There might be a case where there are some unexpected obstacles can cause delays/ road blocks to the renovation process which would cause discrepancies to the provided estimates.

• For example, availability of resources and labor are important to get the work done. In case, the suggested design involves resources which are rarely available, it can cause delays or incomplete renovation.
• Also, if the renovation involves painting the building, the weather conditions need to be taken care. Paining the building in rainy season would be a bad idea and can cause waste of time and money.