Price yield automation and machine learning

min £36 GBP / hour

已关闭

已发布

超过 6 年前

min £36 GBP / hour

Hi, My organisation is an online travel agent. To maximise our sales volume we manually yield our prices by doing the following: 1. Set the price based on a key competitor 2. Monitor the sales volume at that price for one week 3. After one week, start adjusting the price in 5% increments every two days 4. Monitor the sales volume - if unchanged, we keep incrementing the price every two days 5. Once we see a noticeable and consistently decreasing sales volume, we either stabilise or reduce the price until we see the peak volume again When this process is multiplied out by 15,000 hotels it is totally unmanageable. So as long as we run it manually, we are forced to focus on a small set of key hotels. Our objective of this project is to achieve the following: 1. Automate the price adjustment (incrementing until peak sales volume starts decreasing) 2. Learn from each price adjustment to start generalising the kind of price margin we should apply to other hotels (of a similar criteria) 3. Maximise sales volume We are looking to have an algorithm built that would measure itself on the number of hotel transactions achieved. If you don't believe machine learning is the appropriate way to approach this issue - we are happy to discuss other solutions. We are looking for a developer who has expeirence in creating an algorithm of this nature and who can translate their technical know-how to our business problem. This has to be approved by our CFO so after an informal chat, we would require a short, but structured written proposal outlining how you would like to work through the project and the estimated costs at each stage. This is our first data automation / ML project in the business - for the right developer who can produce business results, there will be strong opportunities for a longer-term relationship. Do not reply with a generic template response to this brief - serious enquiries only. Many thanks, Simon

Artificial Intelligence

Data Analytics

Data Mining

Machine Learning (ML)

项目 ID: 15602640

关于此项目

15提案

远程项目

活跃6 年前

想赚点钱吗？

电子邮箱地址

在Freelancer上竞价的好处

设定您的预算和时间范围

为您的工作获得报酬

简要概述您的提案

免费注册和竞标工作

15威客以平均价£41 GBP/小时来参与此工作竞价

@winnow1

Simon, This problem comes under the larger scope of pricing analysis and optimisation area. ML can be applied here but before that we need to do some serious quantitative analysis (price-elasticity curves, optimal price band calculations etc.). For e.g., these analyses would answer questions like 'if we increase the price by x% by what % points would the sales dip?'. This can be automated for 15K hotels in your database. Later ML can come into the picture and help further fine-tune the process (for eg., what are the factors that influence the demand and how much role does pricing have to play etc.) I would like to have a short 30-min call with you to sync our frequencies and expectations, later followed by a proposal. I'm available today before 2pm GMT or anytime from 8AM to 4PM GMT on Mon/Tue.

£50 GBP 在40天之内

5.0

(5条评论)

6.3

@srajbr

Hello, I don't think its the right time to go for Machine Learning right now. I suggest you go for automation script which will manage all the hotels. I will need in more details about each step you mentioned, so that I can design an automation price control system. Regards, Samiran

£40 GBP 在20天之内

5.0

(6条评论)

6.4

@helmot

Hello. I have a Masters degree in AI and have worked on a lot of AI/ML projects for more than 8 years. Also I have worked with eCommerce websites like Amazon, Ebay, ... and have developed some smart tools for sellers. I can show you some demos if you are interested. Thanks, Ivan

£40 GBP 在40天之内

4.9

(9条评论)

5.8

@chrisvwn

Hi Simon, Your organization is trying to automate a price-setting strategy that is time consuming. In addition, you are looking at exploring machine learning to detect and understand trends in the online travel agency's historical pricing and use this to develop a model that maximizes sales volumes. I have experience using R to build exploratory and predictive models to understand data and forecast outcomes so as to help with decision-making. I look forward to working together on this project. Regards, Chris

£40 GBP 在40天之内

4.6

(14条评论)

5.3

@TheAVashe

I've bid for up to 10hrs/week with 45+taxes. If you're really interested for longer-term collaboration, and you're able to make milestone payments, if not always, mostly regularly, I'd be glad to work with you. Coming back to your project: It is my understanding that Steps 1-5 are repeated several times, every week, for every hotel. So, it's just related to hotels (not logistics for now, actually here in our area travel agencies deal just with logistics, so). And, this has to be done separately for 15,000 (possibly more, when automated, which should be the goal obviously). Objective: Optimal Price, Maximum Sales (from data), and automated software (have to discuss about it) What can be done? All the steps 1-5 are possible to be automated. What are the concerns? 1) Do you have any real-time system to get data? Why this is important? Because, quite clearly your automated system has to work on-line (in machine learning jargons), meaning computation won't be off-line several days after data has been collected, but instantaneously, so to speak of, as much as possible, and keep updating itself accordingly. 2) How will you use the algo? What's the platform (Windows/Linux), programming language (R, Python) etc. suitable for you? Is there an alternative? a) Yes, a minimal portable setting can be done that you can carry around in a disk, plug in and work anywhere. b) mathematical formulation, algo as YOU SAID, can be obtained, and you can later get it converted to a standalone software (this may be be part of the LONG-TERM deal) Approach: Of course machine learning should be useful, and a detailed brief shall be done. So, finally being a research student, I'm interested in above stuff. Kindly leave your queries or clarifications even if I'm not online, and shall revert back ASAP. Your brief was well written. Thank you!

£50 GBP 在10天之内

5.0

(3条评论)

3.6

@video556

A proposal has not yet been provided

£38 GBP 在1天之内

0.0

(0条评论)

0.0

@sopholos

Machine learning used to find unknown patterns in huge set of data. As I see in current situation you want track changes entire volume according price changes. So this is not unknown pattern, I does not see that you need machine learning (at least in this project). But, from other hand, you could have different volume of sales because of holidays/weekends/some big meetings (conferences etc). In this case you probably need ML to estimate volume of sales in the future and then compare with real count. So I suggest to start collecting data firstly, then we could check without ML and then with ML. You need collect as many data as you could to use ML properly. For example you need store all data according: Hotels (place, stars, rating etc) Customers (age, sex, count of travels before/in this year etc) Date (holidays in country of hotel/country of customer, meetings, symposiums, exhibitions (near of hotel place) etc) Price (for customer) And all data you think that could affect on volume of sales. Then we will try to find trends and play with ML (I suggest Azure ML as I have big experience with Azure)

£52 GBP 在40天之内

5.0

(2条评论)

0.0

@Saurabhkhanna07

Needs Price Elasticity Model Simulator for projections with demand Cross sell model for growth enhancement Automated, Visualization based tool Relevant Skills and Experience 16 years Analytics experience, primarily in Marketing Usa market experience in financial services and telecom and logistics Programming

£40 GBP 在10天之内

0.0

(0条评论)

0.0

@rasoulnarimani

Hi there, I have more than two years of experience developing machine learning algorithms. In addition, I have a phd in physics and have advanced data analysis skills. I think the first part of the project, automating the price adjustment, does not require machine learning. My understanding is that for this part you need to do dynamic price adjustment while keeping an eye on your competitor(s). It seems that you are using some golden number (5% increments in two days) in your price adjustment process. Before even starting the project I would like to know if you have done any research regarding these values - Here is how I would like to tackle this challenge: Phase 1: Automate your process exactly as it is now, i.g. 5% increments in two days Phase 2: Run A/B tests to optimise these values towards maximising sales Phase 3: Consider adding other variable(s) to the model (Since you are building these algorithms for more than 15000 hotel I suggest to do A/B testing to re-evaluate your re-adjustment process- I am willing to talk you through A/B if needed.) As I said you don't need any ML for any of the above but if you guys have enough data I am willing to work with you to implement ML algorithms for your system. Of course I need to know more about the nature of your data. Hope this provides you some ideas about my background and my problem solving style.

£40 GBP 在40天之内