MHB Form some model to predict a food items probability of you know being sold

[J132456]

Hi so I'm really new to stats, like I know mean, median, mode, and t-testsish and that is it. However, this topic really resonates with me because I know it can have real world applications. So these are my questions.

First off here is my real world problem. At my workplace, we sell food that we make. However, we have to predict how much to make before hand otherwise if we run out, the wait will be too long and people get antsy (Most of the people get mad at me, but I'm just the cashier) Next off, if we make too much food, we end up wasting money. We throw away what we don't sell, except for some food which can last a little longer. It is a loose loose scenarion unless we can predict how much to make with some degree of certainty.

As you can probably tell, I want to form some model to predict a food items probability of you know being sold. Now I'm going back to work soon so I'm starting with these variables. I can base a model on how many people have bought tickets, the pre-sales. However on holidays, there are usually lots more people. In addition, I can base it on weather, because we are out doors, there are less sails when it rains. In addition, certain items are more popular than others. I believe that using these factors, and hopefully others that you guys can bring up, will be enough to start some sort of model to predict all this stuff on.

Next, I was hoping you guys knew of any good free online texbooks I can get my hands on, or even cheap ones I can order online.

Thanks again and I can't wait to hear back from you guys.

 

[Jameson]

Hi so I'm really new to stats, like I know mean, median, mode, and t-testsish and that is it. However, this topic really resonates with me because I know it can have real world applications. So these are my questions.

First off here is my real world problem. At my workplace, we sell food that we make. However, we have to predict how much to make before hand otherwise if we run out, the wait will be too long and people get antsy (Most of the people get mad at me, but I'm just the cashier) Next off, if we make too much food, we end up wasting money. We throw away what we don't sell, except for some food which can last a little longer. It is a loose loose scenarion unless we can predict how much to make with some degree of certainty.

As you can probably tell, I want to form some model to predict a food items probability of you know being sold. Now I'm going back to work soon so I'm starting with these variables. I can base a model on how many people have bought tickets, the pre-sales. However on holidays, there are usually lots more people. In addition, I can base it on weather, because we are out doors, there are less sails when it rains. In addition, certain items are more popular than others. I believe that using these factors, and hopefully others that you guys can bring up, will be enough to start some sort of model to predict all this stuff on.

Next, I was hoping you guys knew of any good free online texbooks I can get my hands on, or even cheap ones I can order online.

Thanks again and I can't wait to hear back from you guys.

Hi piano123,

Welcome to MHB!

This is a really interesting problem as it's directly from the "real world". These are types of problems that I try to think about each day. What you are trying to do is forecast the expected food demand for the next 1-2 days so you can buy appropriate supplies. This is part of a very interesting but unfortunately complicated part of statistics called time series.

My recommendation would be to do some analytics on your sales rather than build a model right now. Do you have a record of all your sales? If you have information with the time it was sold you can analyze it in Excel, looking at sales by time of day and day of the week. Those would be the first two places I'd start. Then add weather, food group, etc.

 

[J132456]

Awesome, this is really helpful. I think I'm going to start with one food item and try to analyze its sales. Do you know of any good textbooks or other sources I could start studying to grasp statistical analysis a little better?

 

[Jameson]

Awesome, this is really helpful. I think I'm going to start with one food item and try to analyze its sales. Do you know of any good textbooks or other sources I could start studying to grasp statistical analysis a little better?

I think this would depend on a few things. Usually students start with a lot of baseline concepts and work their way up slowly. I think that works well if you have a few years but in corporate, business environments I think the goals are a little different. I would recommend doing statistics simultaneously as you learn Excel, Python, or R. These tools are all very standard now in data analytics and will do a lot of things very fast. What type of learning style do you have? What is your main goal for studying stats?

 

[J132456]

So, I do know excel. I can do T-tests, I get p-values. I have some knowledge of standard curves and I have a good basic understanding of integration and differentiation. With this in mind, I know there is so much more to excel than the baseline I have right now. After this summer I'll be a senior and will be taking AP stats. This little summer project is really just for me. I want to see how far I can go in sort of undertanding stats in a business setting and ultimatley I'd like to obtain some model that can accuratley predict the food sales at my job. Since school is winding down I thought it would be a good time to get a jumpstart on challenging myself. I do like other types of math but I see the most utility in having a greater understanding of statistics. That is why I really want to hunker down and learn as much as I can about this topic. So If you know of any good resources that I can start taking a crack at, it would be super appreciated. Also, I plan on coming back here a lot.

Thank you very much for your quick responses and for your helpful insight.