Pytorch for RETS deep learning

Python — incorrectly specified categories?

27 USD

Client's review of cooperation with Sergiy Isakov

Pytorch for RETS deep learning

27 USD

Thank you for the successfully completed project!

5.0

Nikolay Mashkov

Reciprocal Feedback

24 January 2020

Freelancer's review of cooperation with Nikolay Mashkov

Pytorch for RETS deep learning

27 USD

Nikolay Mashkov (Nikolay_168) in the framework of the project "Neuroset on Pytorch for RETS deep learning" set an interesting task and proposed to use for its solution the original approaches related to the development of neuron networks to predict changes in real estate prices.
During the work, the communication with the customer took place at a high professional level and had a positive impact on the implementation of the project as a whole.
We discussed ways of further possible cooperation in this and some other directions.
The customer has a high level of competence and honesty.
Payment is made without delay and in full amount
I recommend all colleagues to respond to the Nikolai projects and you will get not only interesting tasks, but also positive emotions from communication and cooperation.

5.0

Sergiy Isakov

Reciprocal Feedback

24 January 2020

Project translated automatically. Log in or register, to view the original Data in the RETS format. HTTPS://en.wikipedia.org/wiki/Real_Estate_Transaction_Standard Stored in mysql You need to make a pytorch script to predict the price for each of the next 300 weeks but for each type of objects.The object type is the arbitrary combination of RETS fields.Additional entry data: https://www.kaggle.com/federalreserve/interest-rates Important: the forecast for different types of objects should be linked to each other.For example, it is noted that one region begins to go cheaper/early before the other [with a constant timely falsehood] - it is necessary that such connections can be detected and used for forecast.Request format: a set of RETS and datetime parameters, from which you need to make a forecast.(Even though the forecast is made for one type of object, the analysis should take into account all types of objects).Whatever transfer all the data in the command line - the request data can be taken from the mysql by the ID request.Answer: JSON object with a mass of their 300 cents.The script should work at 2 gigabytes of operation.

Update #1 from 9 December 2019

С вопросами и предложениями - пожалуйста звоните на на скайп nicholas3mobile или напишите как вам удобно созвониться.

Proposals 4 Discussions 2

Sergiy Isakov

Winning proposal

10 0

Projects 10
Rating -
Rating 1 263

Budget: 1200 UAH Deadline: 10 days

Nikolai, good day
I analyzed your project and I felt it wasn’t as simple as it seemed at first sight.
You seem to have some hybrid datasets and timelines from which you need to make forecasts in time.
Based on this, predicting the time and amount of costs for the implementation of the project is almost impossible.
Therefore, I can offer you the following cooperation scheme:
Data from the indicated sources
Pre-processing of the source data, if necessary.Analysis of emissions, filling out missed data, etc.Analysis of interconnection, interaction and correlation of data.Selection of independent parameters
- analysis of the stagnation of timelines, trends, timelines, periodicity, abnormal areas, etc.You may need to cluster the timelines.
- construction of the model - can be on a patch or tensorflow
Training, testing and so on.Creating a user interface etc.Maybe something will need to be added and relaxed.
I am ready to take it under the following conditions:
- assign +\20 hours per week (with checking the spent time in a convenient way for you)
The price is $11 per hour.
Payment per week in GBP on the day of payment
Work through the seafood, I pay the commission.
I will send you an invitation on Skype and we can communicate more in detail.
Happiness for you
by Sergey