Excel Test #2

Markov Chains

• Markov Chain: A Markov chain is a stochastic (random) model describing a sequence of possible events in which the probability of the next event depends only on the state attained current event – (e.g., the probability it will rain tomorrow depends on whether it is raining today.) 

• created by Andrey Markov

• example: A restaurant's daily menu forms a Markov chain where tomorrow's dish is solely determined by today's choice, represented by probabilities in a transition matrix.

 

Permutation vs Combination

Permutation

• Permutation is the arrangement of items in which order matters.

• Number of ways of selection and arrangement of items in which Order Matters  

  • arrangement, schedule, order

    n_P_r = n! / (n-r)!

Combination

• Combination is the selection of items in which order does not matters.

• Number of ways of selection of items in which Order does not Matters  

  • group, sample, selection 

    n_C_r = n! / r! (n-r)!

Attenuation vs Amplification

• Attenuation: talks about the process of a signal decreasing. when a brand’s message loses clarity or relevance - it may result from complex messaging, bad targeting, or too many intermediaries.

  • a loyalty program that has too many conditions 

• ex: ex me getting to a person through 10 people is weaker than just getting to them through 1

• Amplification: Amplification is the process of increasing the strength of a signal. - When a message gains reactions and is widely spread, maybe the ad is fear-provoking.

  • marketing goal is to amplify the message through user-generated content to have an impactful big message.

Z-Tests

• Purpose: helps determine if there's a significant difference between a sample mean and a hypothesized population mean, assuming the population variance is known. 

• calculates the one-tailed p-value of a z-test, determining the probability that a sample mean is greater than a given hypothesized population mean. 

Holt-Winters

The Holt-Winters method in Excel is a time series forecasting technique that applies exponential smoothing to data with trend and seasonality components. It helps predict future values based on past data patterns.

ex: on a sunny day people buy more clothing, so to predict what we should stock we should look at past sunny days and the amount that was bought to then be able to predict

What It Does:

1. Accounts for Trend – Adjusts forecasts based on increasing or decreasing trends in the data.

2. Handles Seasonality – Captures recurring patterns (e.g., monthly or quarterly cycles).

3. Smooths Data – Uses weighted averages to reduce noise and make trends clearer.

Price Bundling and solver

finds the best possible solution to a problem by adjusting variable cells to maximize or minimize an objective cell, while adhering to any specified constraints

finding the best prices of different bundles of a product and what pricing leads to people buying the most or the company getting the most revenue

Multiple Regression

analyzes the relationship between one dependent variable and two or more independent variables, helping identify how different factors influence a specific outcome and make predictions or identify trends.

for example - how service importance and quality importance effect number of customers per day