Respuesta :
This is a good inquiry, especially applicable to the study of computers and technology.
It sounds like a prediction problem, in terms of measuring multiple variables calculating their betas and outputting the probability of choosing the following given we have a final the next day.
So let’s start by declaring variables for our model:
Calories: quantitative variable
Consumption time: time to eat the given food also quantitative
Cooking time: quantitative
Baking necessary: categorical 0 = no 1 = yes
Frying necessary: categorical 0 = no 1 = yes
Satisfaction brought: quantitative
Healthy: categorical 0 = no 1 = yes
Now create a data table for implementing our multinominal logistical regression function
I’ll be using 1 serving size for each measure
Fried chicken and onion rings
Cal: 320 (half chicken) + 240 (9 rings) = 560
Consumption time: 10 minutes (avg of sampling data with SD of 2.1 minutes)
Cooking time: 90 minutes + 20 minutes = 110 minutes (based on compiled recipes)
Baking: 0
Frying: 1
Satisfaction(max 10): 9 (-1 for oiliness)
Healthy: 0
Burgers and ice cream
Cal: 354 + 137(1/2 cup) = 491
Consumption time: 8 (mean with sd of 3.2)
Cooking time: 23 minutes (recipe) + 30 minutes (travel time to store and from) = 53 minutes
Baking: 0
Frying: 0
Satisfaction(max 10): 10
Healthy: 0
Pizza and cookies
Cal: 570 (2 slices) + 142 (1 cookie) = 712
Consumption time: 9 minutes (avg of sampling data with SD of 1.3 minutes)
Cooking time: 50 minutes + 72 minutes = 122 minutes (based on compiled recipes)
Baking: 1
Frying: 0
Satisfaction(max 10): 8 (not lactose intolerant friendly)
Healthy: 0
Grilled chicken and vegetables (roasted)
Cal: 162 (chicken breast) + 147 (1 cup) = 309
Consumption time: 9 minutes (avg of sampling data with SD of 1.46 minutes)
Cooking time: 30 minutes + 45 minutes = 75 minutes (based on compiled recipes)
Baking:1
Frying: 0
Satisfaction(max 10): 6 (cause chicken breast is dry and vegetables are nasty)
Healthy: 1
Now to develop our prediction model.
With training and validation data I was outputted the following beta coefficients.
X-int = 2.74
B1: -2.35 with p value of 0.002
B2: -0.72 with p value of 0.038
B3: -0.6 with p value of 0.047
B4: -1.2 with p value of 0.371
B5: -0.81 with p value of 0.016
B6: 2.91 with p value of 0.000
B7: 0.2 with p value of 0.007
Now building our prediction model we combine beta values with their coefficients then take e^x of each of the given options to find the probability of choosing each.
It sounds like a prediction problem, in terms of measuring multiple variables calculating their betas and outputting the probability of choosing the following given we have a final the next day.
So let’s start by declaring variables for our model:
Calories: quantitative variable
Consumption time: time to eat the given food also quantitative
Cooking time: quantitative
Baking necessary: categorical 0 = no 1 = yes
Frying necessary: categorical 0 = no 1 = yes
Satisfaction brought: quantitative
Healthy: categorical 0 = no 1 = yes
Now create a data table for implementing our multinominal logistical regression function
I’ll be using 1 serving size for each measure
Fried chicken and onion rings
Cal: 320 (half chicken) + 240 (9 rings) = 560
Consumption time: 10 minutes (avg of sampling data with SD of 2.1 minutes)
Cooking time: 90 minutes + 20 minutes = 110 minutes (based on compiled recipes)
Baking: 0
Frying: 1
Satisfaction(max 10): 9 (-1 for oiliness)
Healthy: 0
Burgers and ice cream
Cal: 354 + 137(1/2 cup) = 491
Consumption time: 8 (mean with sd of 3.2)
Cooking time: 23 minutes (recipe) + 30 minutes (travel time to store and from) = 53 minutes
Baking: 0
Frying: 0
Satisfaction(max 10): 10
Healthy: 0
Pizza and cookies
Cal: 570 (2 slices) + 142 (1 cookie) = 712
Consumption time: 9 minutes (avg of sampling data with SD of 1.3 minutes)
Cooking time: 50 minutes + 72 minutes = 122 minutes (based on compiled recipes)
Baking: 1
Frying: 0
Satisfaction(max 10): 8 (not lactose intolerant friendly)
Healthy: 0
Grilled chicken and vegetables (roasted)
Cal: 162 (chicken breast) + 147 (1 cup) = 309
Consumption time: 9 minutes (avg of sampling data with SD of 1.46 minutes)
Cooking time: 30 minutes + 45 minutes = 75 minutes (based on compiled recipes)
Baking:1
Frying: 0
Satisfaction(max 10): 6 (cause chicken breast is dry and vegetables are nasty)
Healthy: 1
Now to develop our prediction model.
With training and validation data I was outputted the following beta coefficients.
X-int = 2.74
B1: -2.35 with p value of 0.002
B2: -0.72 with p value of 0.038
B3: -0.6 with p value of 0.047
B4: -1.2 with p value of 0.371
B5: -0.81 with p value of 0.016
B6: 2.91 with p value of 0.000
B7: 0.2 with p value of 0.007
Now building our prediction model we combine beta values with their coefficients then take e^x of each of the given options to find the probability of choosing each.
