toppy007

Sorry todd If you've added your name to the email list it'll all be included in that but I'm not wasting my time with people on this forum who want to pick peoples Gramma apart for fun. .

Hi everyone, firstly just wanted to say thank you too everyone who contributed. I've decided to take a different direction on data collection because I'm not going to get the numbers required. If it was summer I'd have sat out at a lock. bit of a shame in terms of a useable product. But many ways to skin a rabbit. I've progressed in the sense of data collection. And have been using NLP and Fuzzy Parsing to obtain data. I'll now be cleansing the data. And selecting a training set from this. I've been drawn towards integrating a reinforced learning algorithm to help optimise the output. Will update all on the results after Christmas only via the people added an email to the survey. This is because of the stress I seemed to have caused many on the site.via my rubbish Gramma, spelling and subject matter.

Thanks for the input And I yes I think I might reword some of these questions. and I completely agree with you on the variables needed. thank you again

Yes that's correct, relate all your answers to the date it was valued. And 'other' if good for your means of valuation Thank you for taking the time

Okay. Look I'm not trying to rub anyone up the wrong way. My written English is not very good so please don't make fun of it. I don't really understand what the issue is with being a boat owner and asking people for information about the boat they have to help my research project. I'm not forcing anyone into it, and I'm not asking if what I'm doing is pointless. And I don't want anyone to give me what they would regard as private information. Its up to you If you want to know anything about the project. Just ask. More than happy to explain my hypothesis.

Hi Arthur yes that data set is priceless for me. In knowing those figures from those dates I can begin to understand how demand and inflammation effect today's prices. Thanks for filling it out.

I have already had a look at survey monkey and its feature are better, however i was not willing to make a financial commitment. But thanks for the suggestion

Live and let live tony I'm taking onboard advise as I go. I am asking for a contribution knowing that it is of no financial reward. it is up to the community if they wish to support or not and i am OK if people don't want to waste their time. I have selected to keep some mandatory fields because i deem these as key factors in the data's features. I am not able to allow navigation without them using google forms. I do state the questionnaire will take about 15 minutes. Anyways. take it easy

Arthur thanks for the recommendation. I'll add 'others' to the bottom. Just for my own sanity can you explain with an example what questions you struggled to answer. I might be able to reword something. As for the required field's these are generally denoted with '*'! do you recommend this is added to the introduction? Regards

Wow She is an old girl, I've extended the date and add Iron Good luck, my just make the England game seem exciting filling this out before!!!

I'm almost up to 30 data sets So big thank you to everyone who has contributed or tried to contribute. Much appreciated

Ok yeah sure. Thanks Its cool

Thanks Alan I've changed that question to your suggested wording thanks. I'd class Webasto & Eberspacher as electrical diesel heaters which I've got as an option. (Or would you think a better wording was apt?) All answers should relate to the date associated with the value you have stated. It seems you've put yourself through enough pain for one day. unless your a gluten for punishment then you don't need to bother repeating the exercise. tho it would be interesting to have in the data set. thanks Thanks ill add them

Hi thanks for the attempt. Would you mind sharing those details so I can add them in? I'll also add a other section. As I'm sure this will happen again too. Regards

Ok thanks Tam. I'm aiming for 400 data sets so everything is useful even if you had to get creative.

Ok thanks everyone. Will get on the spell check and later. I'll add Berkshire. As long as you answer the questions based on the time the boat was valued! Via purchase or sold or valued (brokerage) I'm happy to collect that data from that date it was done. As for a boat from aboard thanks for trying to fill it out but yes I have kinda set it up for UK canal boats narrow or wide because of the data required to add such features. But once again big thanks. My spelling sucks big time. so apologies

Yeah for sure, ill be processing the data before using it, looking for these types of things too. I'm not sure the scope will consider this but you can continuously train these types of models. so for example it would be learning from a larger about of data as sales and time went on, strengthening its estimation predictions. but like i said baby steps and a minimal product is important to the project. did you complete the questionnaire or also not because its a waste of time?

Hi Everyone I am on the hunt for data from narrow boaters, I created a discussion on the student project for a narrow boat value estimation program. I've had a lot of comments about the project. I just wanted to set this topic up so people could access the link to the questionnaire more easily. <iframe src="https://docs.google.com/forms/d/e/1FAIpQLSfa6nIrVYM8PhyEX1qUBbhw0BRmmuN8VsOJ0bDljV_RJZA3yQ/viewform?embedded=true" width="700" height="520" frameborder="0" marginheight="0" marginwidth="0">Loading…</iframe> The link is https://forms.gle/dHw5J6dihtVws3K16 I've had 9 people take part and would like to get to 400 use this topic to make suggestions on it, if you don't understand a section or feel things can be worded better. Also let me know if you can recommend anywhere to get people to take part. Kind Regards Chris and if you want to discuss any aspects of the idea please have a look at the narrowboat valuation program discussion. Please

Blimming heck Allan youve worn me down! congratulations! I canny be assed with answering this question! Pub and ale calling on a Friday and the snookers on. Do us a favour and just contribute, or don't but honestly once the program is finish I'm more than happy to publish the results, methodology and code for you to debate. I half think you are trying to push my buttons 🤣

Yes you are correct. Collecting training data that is accurate is one of the most important details. And I don't no the outcomes will be valid if the training data is made up! But as for a proof concept maybe it would work. Once trained the program should be able to consider The owners own evaluation in aspects of the boat. Comparing things like age and service history against the owners personal opinion. did you end up filling the questionnaire out? would you like me send you some more information on ML? you sound like you've brought and sold alot of boat so your data would be great to include I think this can definitely be incorporated into a valuation quite simply to. For example sale prices where boats are sold needing finance and others sold without. As well as quick sale data! Date advertised to date sale agreed with demand and market availability considered. Im just not including this data at the current point in time as it is currently MVP.

I think this can definitely be incorporated into a valuation quite simply to. For example sale prices where boats are sold needing finance and others sold without. As well as quick sale data! Date advertised to date sale agreed with demand and market availability considered. Im just not including this data at the current point in time as it is currently MVP.

Ahahaha typo 'data'! Apologies

Hi Tony Machine learning algorithms are able to understand these features, they learn for themselves from the date they are given. the aim is to feed the ML algorithm data such as this! (location, condition, mooring) for it to learn and creates its own model of which to generate a figure. I intend to use a supervised algorithm this means from the training data ill also be telling the ML algorithm the value of the boat from that set. The catch is the quality of the date and amount of data. (feature enhancing can also be applied to the date to aid this) hence the more people who contribute to the data the stronger the results tend to be. Is there anything else you'd like me to explain about it?

OK firstly, thumbs up for a round one's attempt and sorry for the miss leading question. I am attempting to ask boat owners(current or previous) for the date they last got a value(price) associated with the boat they own/owned and what feature it had at that date. Historical data and demand at the time are very important to features for the computer program to learn from. After the date question. value and where the figure came from questions are asked. https://forms.gle/Um3oQMR9mgvRDij79 feel free to give it another go as no information if useless.

Dear Tony Yes! I understand that a value based on historical trends and current demand is not going to decide the sale price of a narrowboat. but it might help buyers better understand asking prices and protect them from paying over the odds if they are less experienced. It might also help a seller, not sell something cheaply if they were getting pushed into a sale. It may also simplify the search process which can be time consuming, one could enter a budget and then balance the feature's they wanted. I guess this program could be seen as an optimization problem. The results should help boaters and prospective boaters simply understand current market value.