Brainscape_all_20190317 Flashcards
1
Q
Overnormalization
A
d
2
Q
1 Byte
A
0 to 255 (8 bit)
3
Q
SQL Lock
A
J
4
Q
Entity
A
Tables
5
Q
Scale (Data)
A
Number of decimal places (123.123 = 3)
6
Q
Parse()
A
Convert strings to dates.
7
Q
Cache
A
d
8
Q
∅
A
Empty Set
9
Q
Overfitting
A
d
10
Q
Foreign Key
A
Links to the primary key of another table.
11
Q
Referential Integrity
A
(Order for employee that doesn’t exist)
12
Q
Superlative
A
An exaggerated or hyperbolical expression of praise.
13
Q
Denormalization
A
SAMS vs RAD
14
Q
Dimension Table
A
d
15
Q
Floating Point Number
A
Efficient, but lose accuracy.
16
Q
CREATE SEQUENCE
A
D
17
Q
Ridge Regression
A
f
18
Q
1st Normal Form
A
No repeating columns.
19
Q
Table Scan
A
Scans entire table contents.
20
Q
Try_Cast()
A
Same as Cast, but generates nulls for errors. (Cast/Convert/Parse)
21
Q
Accuracy
A
d
22
Q
clf
A
Classifier
23
Q
Candidate Key
A
Possible primary keys.
24
Q
EDH
A
Enterprise Data Hub
25
Don't 1
Meddle
26
Clustered Index
Pyramid
27
API
Application Program Interface.
28
Crossvalidation
Randomly breaks data into parts (80/20 training/testing).
29
Hyperbole
Exaggerated statements or claims not meant to be taken literally.
30
Uniqueness Constraint
Generally covered by primary key, but could also be valid for surrogate keys.
31
Windowed Functions Syntax
Function() Over (Partition By x,y,z Order By a,b,c,x,y,z)
32
Data Integrity
d
33
Attributes
Columns
34
.iloc
d
35
Use full object name.
Servicer/Database/Schema/Object. Improves performance because SQL doesn't have to figure out.
36
Checkpoint
d
37
Forms
Every attribute must provide a fact about the key, the whole key, and nothing but they key, so help me Codd.
38
Density(Index)
Uniqueness
39
K-fold Cross Validation
d
40
2 Byte
+-30,000 (smallint)
41
Don't 3
Accept, reject, turn away.
42
.head()
Number of rows to select (pandas dataframe).
43
Sword of Damocles
Representing an impending ever-present threat. Especially to someone in a position of power.
44
8? Byte
bigint (64 bit)
45
GUID
Global Unique ID
46
Create Procedure
CREATE PROCEDURE schema.name AS ...; ...; go EXECUTE schema.name (@=x)
47
Overtesting
d
48
3rd Normal Form
All fields must only depend on a key. (3NF)
49
Recall
d
50
Ontology
The study of existence, reality, being."What can be said to exist?""What is a thing?""Into what categories, if any, can we sort existing things?""What are the meanings of being?"
51
Universal No Lock
SET TRANSACTION LEVEL READ COMMITED
52
Fixed Precision
Decimal/Numeric. More costly. More precise/accurate than float.
53
Type 1 error
False positive
54
Confusion Matrix
d
55
Overfitting
Overly complex model fits original data very closely, but is poor at making predictions from other data sets, because irrelevant connections are made.
56
Normalization
d
57
Selectivity(Index)
By how much does the field reduce the values to be searched.
58
Acclimate
Become accustomed to a new climate or to new conditions.
59
Precision
d
60
Primary Key
Unique value per row. Can be made up of multiple columns (composite) key. Functions as a target which a foreign key can reference.
61
Random Seed
f
62
Obsequious
Excessively obedient or attentive. Sycophantic.
63
Depth(Index)
d
64
Data Type Command
Type(x)
65
Exists
Returns True/False
66
⋃
Union
67
Epistemology
Relating to the theory of knowledge. What does it mean to know something. How do we know what we know.
68
Entity Integrity
Looks across columns. Uniqueness.
69
Dependency
Columns retrievable by key? In order to lookup a value, you have to know the key first.
70
OLAP Cube
Pre-groups to save time.
71
Transaction(db)
A sequence of database operations that satisfies the ACID properties (can be perceived as a single logical operation on the data).
72
ODS
Operational Data Store
73
Leaf Level (Index)
All the data.
74
Overtraining
d
75
Probability Axioms
P(A) ≥ 0P(Ω) = 1If P(A⋂B) = ∅ THEN P(A⋃B) = P(A)+ P(B)
76
Fact Table
d
77
String Values
Larger because there are way more than 9(10?) values per character.
78
Connect to Python
import pyodbcconn = pyodbc.connect('Driver={SQL Server};' 'Server=he2qntvpsql382;' 'Database=MHASAMS01PDB;' 'Trusted\_Connection=yes;')cursor = conn.cursor()cursor.execute('SELECT Organization FROM MHASAMS01PDB.ra.OrgLookup')for row in cursor: print(row)
79
Help Command
Help(x)
80
Duplication
Attribute unnecessarily repeated. (SAMS and RAD duplication.)
81
Json
Identifies data types auto
82
OLAP
Online Analytical ProcessingOptimized for read only. Good for analytics.
83
Disjoint
P(A⋂B) = ∅
84
Batch Gradient Descent
d
85
SVD Decomposition
d
86
obsequious
obedient or attentive to an excessive or servile degree."they were served by obsequious waiters"
87
Schema
Groupings of objects (tables). Can grants permissions based on schemas.
88
Who's causing you harm?
You are.
89
Default Value For Data Integrity
k today(), getdate()
90
EDW
Enterprise Data Warehouse
91
OLTP
Online Transaction ProcessingTypically facilitate and manage transaction-oriented (UPDATE, INSERT, DELETE, ALTER) applications vs analytical (OLAP).
92
⋂
Intersection
93
2nd Normal Form
Non-key columns shouldn't be dependent on a part of the primary key.
94
Validation
Find out which model is the best.
95
Persisted
SQL Function
96
SQL View
USE Database; go CREATE VIEW schema.viewname AS Select...;
97
ACID
Atomicity- Guarantees that each transaction is treated as a single "unit", which either succeeds completely, or fails completely.Consistency- Ensures that a transaction can only bring the database from one valid state to another (transaction follows constraints, cascades, triggers, etc...)Isolation- Running transactions simultaneously produces the same outcome as running them sequentially.Durability- Once a transaction has been committed, it will remain committed. (Recorded in non-volatile memory).
98
Scoping
D
99
Domain Integrity
Integrity within column (consistent data type, nullable, add rules(1:10), prior to x date)
100
Referential Enforcement
Cascade :Deletes all children when parent is deleted.Set Default: Catch all value when parent gets deleted.set null
101
Surrogate Keys
Alternate primary keys that can be used to check the validity of primary key.
102
Why not index everything?
Indexes have to be updated every time table changes.
103
Equivocate
Use ambiguous language so as to conceal the truth or avoid committing oneself.“Not that we are aware of,” she equivocated"
104
Not
Separate
105
Variable scope, block scope
D
106
4? Byte
+- 2,000,000,000 (int) (32 bit)
107
Insert Into With Select
Insert IntoSelect \* FROMWHERE
108
Gradient
Slope
109
Dot Product
Inner Product
Single value.
110
Orthogonal
Perpendicular
dot product = 0
111
Staging Tables
Raw data is put into table, then processed from that table. Faster than directly processing data source.
112
Measure
Measures provide information on a given process such as the process of selling something. For example, the measured value 15 may represent a number of items sold.
113
Fact Table
Most of the time, Fact tables are design around an event that happened within your company, for example the event of a product being sold.
114
Dimension Table
When talking about a data warehouse dimension we are describing different aspects of the measured event; what product was sold? What customer bought it, what date did it happen, and on what sales order and line item did it occur.
(Data)
115
IDE
Integrated Development Environment
Includes editor, debugger, automation.
VSStudio, VSCode etc...
Prog
116
Insert
array\_name.insert(loc,value)
prog
117
Append
array\_name.append()
118
gini attribute
measures its impurity:
a node is “pure” (gini=0) if all training instances it applies to belong to the same
class.
Green Node
![]()
119
feature
120
Cart Training Algorithm
data
121
122
123
