2DA3 Midterm Review Cards (off practice problems)

Decision-making process

The structured sequence of identifying and defining the problem, determining the set of alternatives, and evaluating those alternatives before choosing one. In the practice key, carrying out the choice is the odd one out.

Strategic decision

A long-term, high-level choice that sets the direction of the whole organization, for example an online-only retailer opening physical stores across the city to complement its web presence.

Tactical decision

A medium-term choice that turns the big strategy into plans for using resources, such as splitting next year's budget across marketing channels.

Operational decision

A short-term, day-to-day choice about running routine activities, such as setting this week's staff schedule.

Descriptive analytics

Techniques that summarize past data to show what has already happened, using tables, charts and summary statistics.

Predictive analytics

Techniques that use historical data and patterns to forecast what is likely to happen next.

Prescriptive analytics

Techniques that take input data and yield a best course of action. Linear programming is the classic example.

Scatter chart

A graphical presentation of the relationship between two quantitative variables, with one variable on each axis and every observation shown as a dot.

Stacked-column chart

A chart that splits each category's column into segments by a second category, such as counts of males and females within each letter grade from A to F.

Pie chart

A circular chart that shows each category's share of a whole as a slice, best for a few parts that add up to 100%.

Heat map

A two-dimensional graphical representation of data that uses different shades of color to indicate magnitude.

Histogram

A chart of touching bars showing how the values of one quantitative variable are distributed across numeric ranges (bins).

Data dashboard

A visualization tool that updates in real time and presents multiple outputs together in a single view.

Card visual

A Power BI visual that displays one headline number, such as total sales, total quantity sold or total number of orders.

Filled map

A Power BI geographic visual that shades regions by their value, used to show sales by region.

Donut chart

A ring-shaped cousin of the pie that shows each category's share of a total, such as sales by deal size.

Bar chart

A visual that compares a value across categories using the length of rectangles, such as sales by product line.

Slicer

An on-report filter control that lets the viewer click values (for example Status or deal size) to narrow what every visual on the page shows.

Conditional formatting

A rule-based setting that changes colors based on the data, such as the background color that highlights the Buy activity in a Power BI report. It is set up from the build pane of the visual, not from the general formatting effects.

Report formatting fundamentals

The design habits a clean report must follow: one consistent color scheme, borders, uniform text size and appropriate titles on every visual.

COUNTIF

Excel function that counts the cells in a range that meet a single criterion. The criterion goes after the range as quoted text with its operator, for example greater than 3 written as >3, to count orders with more than 3 items.

AGGREGATE

Excel function you reach for when rows are hidden by a filter and you still want an average or sum that ignores them. A plain average would count the hidden rows too.

Dollar-sign cell anchoring

A dollar sign before a column letter freezes the column, one before a row number freezes the row, and both freeze the whole reference when copied. For example A$9 freezes only the row.

IF with AND

Formula pattern that returns 1 only when every test is true and 0 otherwise, such as payment method is Visa and total cost is above $100.

IF with OR

Formula pattern that returns 1 when at least one of the tests is true, even if the others fail, and 0 only when all of them fail.

Pivot table

Excel summary tool that rearranges a dataset by dragging fields into rows, columns and values, such as total cost by region down the side and payment method across the top.

One-way data table

What-if tool that shows how the changing values of a single input affect an output of interest, such as monthly car payment under different interest rates.

Two-way data table

What-if tool that varies two inputs at once, such as interest rate and down payment. One input's values run across the top row and the other's run down the left column, and each input needs its own input cell box.

Data table setup rules

The corner cell must link to the output formula being studied (for example =$B$8), and the input cell boxes must point at the original input cells that get swapped out. Typing results into them or leaving them empty breaks the setup.

Solver add-in

Excel tool that finds the best answer by setting a target cell to maximize or minimize, choosing the cells it can change, and respecting the constraints. Its Answer report must be included with a neat model.

Decision variables

The quantities the modeler controls and wants to determine, such as the dollars to invest in each of five companies.

Objective function

The expression that defines the quantity to be maximized or minimized, such as the expected return for a manager choosing among stocks and bonds.

Constraints

Restrictions on the variables written as

Nonnegativity constraints

Conditions that guarantee the solution only contains values of zero or above for the variables you are solving for.

Feasible solution

A set of variable values that meets every restriction at the same time. Meeting only a few is not enough, for example (3,1) breaks 5x1 + 2x2

Infeasibility

The situation in which no set of values meets all the restrictions at once, so the problem has no answer.

Unboundedness

The situation in which the objective can be improved forever while still meeting all restrictions. In real problems it signals the formulation is improper, such as a missing restriction.

Binding constraint

A restriction that holds as an equality at the optimal solution, meaning zero slack and the resource is fully used up. Geometrically, these lines intersect at the optimal point.

Extreme point

A corner of the feasible region on the graph. In the graphical method you check the objective value at each one and pick the highest (max) or lowest (min) to find the optimal solution.

Linear programming assumptions

The three conditions a model needs to be appropriate: contributions and resource use scale directly with each variable's value (proportionality), totals are sums of the parts (additivity), and variables may take fractional values (divisibility).