Please enable JavaScript.
Coggle requires JavaScript to display documents.
Science & Tech XXI Century - 1° Register (Charts (Frequency Table…
Science & Tech XXI Century - 1° Register
Learning Process
Supervised Learning
There is a training process; a specific classification criterion, and labeled results as 'right' or 'wrong'
Unsupervised Learning
The process of learning without training labels; there are no correct answers, just interesting patterns
Variables
Qualitative
less precision an accuracy
measured with observation, surveys, and interviews
results are based on descriptions and quotes
Ordinals
No numerical values in hierarchy
Nominals
Unordered categories, labels
Independent
manipulated characteristic that causes the studied phenomenon
Quantitative
Accurate and exact data
Easy to generalize (sample -> population)
Measured with surveys, instruments and questionnaires
Dependent
/result generated in the manipulation of the variable X
Charts
Frequency Table
shows qualitative results according to
their frequency
Combined Frequency Table
used for categories with multiple variables
Bar Charts
showing Qualitative data results in terms of how frequent
Relative Frequency Table
frequency + calculated percentage
Pie Charts
shows data as portions of a pie depending on their percentage
Histograms
bar charts squished together to show continous data
Binning
Manipulating the data on a graph to make it show or hide certain tendency
Distribution Curve of a Graph
Normal
the mean, mode, and median of the distribution are equal
symmetric
mean = central tendency
Skewness
used to "correct" data when it's not normally distributed
skewed data can have a greater tendency to the right or left
depending on the tail we know if it's right or left skewed
Data Transformation
we can transform data to make it easier to visualize
we can use square roots, logarithmic formula or squaring to transform the data
Scaling
transforms the values to another range
magnifies or reduces the range
methods: mean centering, limit scaling, standarization
ALGORITHMS
Decision Trees
clasification trees: yes/no
regression trees: continous data type
flowcharts used to predict an outcome
Random Forest
extension of multiple decision trees
K-Nearest Neighbour
stores data and classifies the new data based on similarities
Artificial Neural Network
simulates neurons