Grouping (グルーピング) is the processing that gathers data by similar traits and divides it into several groups.
All members
Frequent visitors who spend a lot
Rare visitors who spend a lot
Members who spend little
Group customers by spending and visit frequency, or products by how they sell, and the traits of each group become easy to compare. People can set the criteria themselves, or the data can be gathered automatically according to how close the values are.
AnalogyLike sorting a messy stack of books onto shelves for fiction, cooking, and travel. Once the shelves exist, you can see at a glance which subject dominates.
Pattern discovery (パターン発見) is the processing that finds regularities and trends that recur in accumulated data.
Collect records
Look for links
Act on the findings
It moves the two displays next to each other and tries a set price.
3 / 3
It probes how items relate and how they shift over time, then feeds the results into product selection and promotions. The technique of digging such regularities out of large volumes of data is called data mining (データマイニング).
AnalogyLike a store manager skimming register records and noticing that prepared foods sell first on Friday evenings. You read a trend that no single record states.
How to remember"Mining" means digging for ore: picture striking a rich vein in a mountain of data.
Exam points
ExamThe classic case of data mining (データマイニング) is analyzing which products are bought together. Totaling fixed fields does not count.
Mix-upA data warehouse (データウェアハウス) is a store that holds data for analysis, which makes it something quite different from data mining, a technique for finding regularities.
Optimization (最適化) is the search for the option that brings the value of the objective (目的) to its best while honoring the constraints (制約条件).
Set the objective
List the constraints
Compare candidates
Adopt the best option
Among the combinations within budget, it adopts the one with the largest profit.
4 / 4
Objectives include minimizing cost or maximizing profit, while constraints include budget, time, and the amount of materials. In problems that must meet supply and demand at once, such as shipping between several sites, filling the cheapest routes first is not always best, so compare the total for each combination. When the future is hard to read, an approach such as the maximin strategy (マクシミン戦略) decides by comparing the worst outcome of each option.
AnalogyLike planning the most satisfying trip you can on a limited budget and number of days. You cannot visit everywhere, so you compare combinations and choose.
How to rememberEnglish "optimize" comes from the Latin optimus ("best"), which points the same way as the kanji in 最適 ("most suitable").
Exam points
ExamWith several materials, output is capped by whichever runs out first. Work out how many units each material allows and take the smallest figure.
ExamTo maximize profit in limited working time, rank products by profit per unit of working time, not per item, and give each its hours in that order up to its production ceiling.
ExamThe maximin strategy (マクシミン戦略) compares each option's lowest profit and picks the option whose lowest profit is highest.
Inventory management (在庫管理) is the work of keeping goods and parts at a level with neither shortages nor excess.
Stock level
What happens
Too low
Stockouts cost sales and production opportunities
Right level
Demand can be met and extra costs stay down
Too high
Storage costs rise and cash sits idle in the form of goods
A stockout (欠品) means the goods are not there when needed. ABC analysis (ABC分析) is one method of focused control: it classes items in descending order of value and manages the top group most closely. Assigning stock to received orders is called allocation (引当).
AnalogyLike household staples. Running out is a nuisance, but overbuy and you run short of space and things pass their use-by dates.
Exam points
ExamA Pareto chart (パレート図) overlays a cumulative-percentage line on bars sorted from largest to smallest, and it narrows focused control down to the items with the largest share of value.
ExamWhen allocatable stock (引当可能在庫) is taken to mean the stock that can be allocated, ordered goods not yet received do not count. Start from the stock at the beginning of the tally, add later receipts, and subtract the quantity already ordered by customers.
ExamWhen a moving average (移動平均, the mean over a recent fixed period) forecasts next month's sales and output is set from the gap against stock, only actual sales are averaged. Stock uses the figure at the end of the current month.
An ordering method (発注方式) is a rule that sets when and how much to reorder, with the fixed-quantity ordering method (定量発注方式) and the fixed-period ordering method (定期発注方式) as the main examples.
Fixed-quantity ordering method
Fixed-period ordering method
Order trigger
When stock falls to a set level
When the set cycle comes round
Quantity per order
A preset amount
An amount calculated each time
What varies
Order timing
Order quantity
The stock level that triggers an order under the fixed-quantity method is the reorder point (発注点). Either method may also allow for safety stock (安全在庫), extra units held against swings in demand and late deliveries.
AnalogyLike a house rule for restocking milk in the fridge. Decide when and how much to buy, and you top up before it runs out, with no second-guessing.
How to rememberRead the kanji: 定量 fixes the 量 (quantity), and 定期 fixes the 期 (period).
Exam points
ExamWith steady demand, the best order size under the fixed-quantity ordering method (定量発注方式) is the one that minimizes ordering costs plus storage costs combined. Larger orders lower the first and raise the second.
ExamThe fixed-period ordering method (定期発注方式) sets the order quantity as (order interval (発注間隔) + lead time (調達期間)) × daily usage + safety stock (安全在庫) − current stock − outstanding orders (発注残). Lead time runs from placing an order to delivery, and outstanding orders are goods ordered but not yet received.
ExamTo get a parts order from a production plan, take the usage for the period this delivery must cover and subtract the stock expected to remain just before it arrives. If safety stock is included, add it on.
Credit management (与信管理) is the work of assessing whether customers buying on deferred terms can pay and setting a credit limit (与信限度額), the ceiling on uncollected amounts, to prevent bad debts (貸倒れ).
Item
Example amount
Credit limit (ceiling on uncollected amounts)
¥10,000,000
Uncollected sales
¥3,000,000
Promissory notes (written promises to pay on a set date) not yet at maturity (that date), if the terms treat them as uncollected
¥1,000,000
Orders received (if the terms include them)
¥1,500,000
Portion of the limit still usable (available credit)
¥4,500,000
A bad debt (貸倒れ) arises when payment for a sale can no longer be collected. The part of the limit still usable is the available credit (与信の余力). Payment records then feed into revising the limit or changing the trading terms.
AnalogyLike the limit on a credit card. Spending shrinks what is left, and repaying restores it.
How to rememberThe kanji in 与信 literally mean "give trust". Allowing a customer to pay later is exactly that.
Exam points
ExamAvailable credit (与信の余力) is the credit limit (与信限度額) minus the uncollected balance. If the terms count orders received as credit and do not treat unmatured promissory notes (手形) as collected, add both to the balance.
The technique of digging regularities out of large volumes of data is data mining (データマイニング).
Optimization (最適化) sorts out the objective and constraints first, then compares candidates.
Inventory management (在庫管理) aims to avoid the lost opportunities of stockouts and the storage and cash burden of excess.
Under the fixed-period ordering method (定期発注方式), the order quantity is the usage for the period the order must cover plus safety stock (安全在庫), minus stock on hand and outstanding orders.
Available credit (与信の余力) equals the credit limit minus the uncollected balance.