Exam1 Flashcards
1
Q
What year was the term AI coined?
A
1955, Darthmouth
2
Q
What is the definition of AI
A
Any system that exhibits behavior that could be interpreted as human intelligence
3
Q
Weak AI is also called
A
Narrow AI
4
Q
____ AI is good for systems that have predefined patterns to eliminate impossible options
A
Planning
4
Q
Strong AI is also called
A
General AI
4
Q
Definition of Weak AI
A
model that is confined to a narrow task
5
Q
What are some examples of weak AI tasks
A
Language to text processing; picture sorting
6
Q
Siri is an example of a weak or strong AI?
A
weak
7
Q
Definition of strong AI
A
the machine displays all person-like behavior that you’d expect from an artificial human (emotions, humor, etc)
8
Q
What was an early name for the nodes in neural networks?
A
Perceptrons (Rosenblatt at Cornell)
9
Q
When did the term “Deep Learning” become popular?
A
1990s
9
Q
Reasons that machine learning has accelerated
A
Availability of data
Moore’s Law
IoT
Automated SW coding (sensors and controllers)
10
Q
Training and test data is
A
labelled
11
Q
3 categories of supervised learning
A
- Binary classification
- Multiclass classification
- Regression analysis
12
Q
If you have massive amounts of unlabeled data, ____ algorithm could be a good choice
A
k-means clustering
13
Q
Bagging, boosting, and stacking are examples of
A
ensemble modeling
14
Q
Definition of bagging
A
create several different version of the ML algorithm in parallel (like decision trees with different roof notes), and compare results, average out
14
Q
Definition of Boosting
A
Use several different ML algorithms in sequence to boost accuracy of results (model 2 learns from model 1 etc)
15
Q
Definition of Stacking
A
Use several different ML algorithms to boost accuracy (ex. k-NN on top of Naive Bayes)
16
Q
For abstract reasoning, a _____ system reasoning may be best
A
symbolic
17
Q
Definition of bias
A
gap between predicted value and actual outcome
18
Q
Definition of variance
A
how scattered predicted values are +/- of actual outcome
19
Q
What is the Turing test
A
Can the machine fool a human into thinking it’s a human if it’s behind a wall?
20
Q
Big data is
A
unstructured data
21
One challenge of using AI for predictions is that AI uses _____ data
Historical (ex how would an AI model fall out of an unanticipated large event like Covid?)
22
One of the reason AI didn't take off in the 60s and 70s was
limits of technological maturity (memory space, computational power)
23
When building an AI model, keep _____ in mind
the end goal in mind: who will use this model any why
24
raw data is
data collected in it's original form, prior to any processing or adjustments
25
3 types of Data analytics
- Descriptive
- Predictive
- Prescriptive
26
Difference between predictive and prescriptive models
Predictive just predict the future (forecasts, etc), prescriptive change the future (control, optimization, etc)
27
Examples of types of data
- numeric vs non-numeric
- categorical data (ex fault or no-fault)
- structured vs unstructured
- temporal, spatial, spatio-temporal
- experimental vs operational
28
Experimental data differs from operational data critically in that ___
experimental data will isolate a single (or few) variables from other variables, while operational data will have a much more impact from the surrounding environment (which was not controlled)
29
Definition of Big Data
data that challenges the current capabilities of a single computing unit
30
What types of data would we encounter in energy systems
- metered data
- sub-metering
- communications
- measured data
- data storage
31
What does CRISP-DM stand for
Cross industry standard process for data mining
32
An input is also sometimes referred to as __
an instance
33
Definition of Data Analytics
the science of analyzing raw data to draw insight, and make conclusions from that data
34
Linear data cleaning workflow
1. Access Data
2. Detect Duty Cycles
3. Remove Outliers
4. Sanitize Gaps
5. Check Process Limits
6. Analyze data...
35
In univariate stats, variance =
std_dev^2
36
Covariance is...
the variance between 2 variables
37
Positive Covariance: variable A increases as variable B
increases
38
In weak covariance, there is...
no apparent linear statistical dependence between the 2 variables
39
Negative covariance, each variable "varies" ....
inversely to the other variable
40
Unlike covariance, correlation is...
- normalized to -1 to +1
- unitless
41
Correlation of A&B =
covariance of A&B / (std.devA*std.devB)
42
____ increase modeling risk
outliers
43
Outliers are:
data points that are significantly different from the rest of the data set
44
What is the simplest outlier detection technique for univariate samples
Z-score, where Z is the standardized equivalent of the data value = (x-x_mean)/std.dev
45
MCD stands for
Minimum Covariance Detection
46
MCD can be used to
remove outliers from multivariate samples
(minimum covariance determinant)
47
Definition of imputation
The process of identifying missing data, then creating a substitute
48
Why is it important to impute data sets
- missing data is generally not allowed in training data sets
- throwing out entire data points could throw out useful data
- statistical techniques could be biased by missing data
49
The covariance and correlation matrices are
symmetric
50
A typical Z score for outlier cutoff would be
3 ( = 3 std dev away from mean)
51
Imputed data inherently introduces
Bias into subsequent modeling
52
What are the 2 options to deal with missing data
1. throw it out
2. fill in the gap
53
What are some ways to impute?
- simple statistics (use mean, median, a constant)
- Multivariate imputation with bayesian stats
- k-nearest neighbor imputation
54
What are some initial questions to ask when prepping data for ML
- does the data include info that can predict the target?
- does the granularity of the training and prediction match?
- is there labeled data?
- is the data accurate? Do you know where it came from?
- is it easily accessible and readable?
- are the missing values a small percentage of the fields of interest?
55
Definition of an algorithm (comp sci)
a sequence of explicit instructions which perform a specific task
56
_____ analysis is used to simplify complexity analysis
asymptotic
57
_____ is a subset of AI
Machine Learning
58
Definition of Machine Learning
the study and usage of both algorithms and statistical models, which computer systems use, without explicit instructions, to learn how to perform specific taks
59
____ is a subset of Machine Learning
Deep Learning
60
Machine Learning applies the fields of
Comp Sci; Optimization; Statistics
61
Unsupervised ML models can be used for
Clustering
62
Labeled data is data which ___
has an associated category assigned to a specific set of features in the data set
63
In hard clustering, each data point...
belongs to only 1 cluster
64
What clustering techniques are examples of hard clustering
k-means, hierarchical
65
Guassian Mixture Modeling is an example of
soft clustering
66
What are some applications of clustering?
- exploratory data analysis
- dimensional (feature) reduction
- image segmentation
- anomaly detection
- data mining
67
Formula for euclidean distance between 2 pts with 2 features
d = sqrt( ( x1 - x2)^2 + (y1-y2)^2 )
68
In K-means clustering, a centroid is...
the arithmetic mean of the points in each dimension
69
Hierarchical clustering can be preferrable over k-means when dealing with
a smaller amount of data
70
What are some convergence criteria you could set for k-means
- % reduction drop of SSE
-Hard stop limit to avoid infinite iteration and/or a known goal
71
Hierarchical clustering creates a _____
dendrogram
72
Gaussian Mixture Modeling is a
probabilistic technique
73
In GMM, the center of the cluster is the
arithmetic mean
74
A model with overfitting is
too complex, maybe has too many predictors
75
You could have ovefitting when
the model is more complex than the data
76
overfitting is ____ common than underfitting with AI models
more
77
What are some applications of classification?
- fault detection
- predictive maintenance
- speech recognition
78
Classification error is quantified by a
loss function
79
What is the formula for inverse distance weighting
w_i = (1/dist_i)/(sum(1 to k)of (1/dist_i))
80
Euclidean distance in 2D is the same as
formula for the hypotenuse of a right triange
81
What are some advantages of k-NN?
- simple algorithm, with flexible options (distance calc method, # of k)
- considered a benchmark for other classification methods
82
What are some disadvantages of K-NN
- sensitive to outliers and erroneous labels
- memory intensive with larger k, pts, and features (giant distance matrices)
83
Resubstitution loss is...
the error just on the training set
84
What are advantages to decision trees?
- can handle non-linear responses
- excellent with categorical variables
- easy to understand for a small number of features
- once you build the model, classification of new data is computationally quick since it is just binary decisions
85
Disadvantages of decision trees
- struggles with a large number of features with smaller data size
- difficult to understand for a large number of features
86
Naive Bayes is a ____ classification technique
probabilistic
87
What are the 3 AI for energy transition principles
- Governing (Risk Management, Standards, Responsibility)
- Designing (Automation, Sustainability, Design)
- Enabling (Data, Incentives, Education)
88
How much investment does BNEF expect to need for a net-zero scenario
between 92 and 173 trillion by 2050
89
What are the 4 main fields where AI could be used in Energy Systems
- Renewable power gen. and demand forecasting
- Grid optimization and operation
- Management of energy demand and DER
- Materials discovery and innovation
90
K-means clustering has an inherent risk that the initial clusters converge....
to a local minimum, rather than global minimum SSE
91
Is K-means sensitive to outliers?
yes
92
Which has a higher time complexity, k-means or hierarchical clustering?
hierarchical
93
