Page 144 - Artificial Intellegence_v2.0_Class_12
P. 144

.    hat do you mean by time series decomposition
           Ans.    ime series decomposition is a fundamental step in time series analysis because it helps in understanding the
                  different contributing factors in the data and can aid in making forecasts or predictions. It considers a series as a
                  combination of level, trend, seasonality and noise components.

              .    ame the components of the time series decomposition.
           Ans.   he components of the time series decomposition are as follo s
                   Level:  he average of the series.
                   Trend: Any increasing or decreasing value in the series.
                   Seasonality: Any repeating short term cycle in the series.
                   Noise: Any random variation in the series.

              .    hat is      algorithm and ho  does it determine the category for a ne  instance
           Ans.       earest  eighbour algorithm is one of the simplest  upervised  earning based  achine  earning algorithms.
                   he      algorithm assumes similarity bet een the ne  case and the e isting cases and assigns the ne  instance
                  to the category that matches the e isting cases the most closely.  his algorithm can be applied to both classification
                  and regression.
        B.   Long answer type questions.

              .     plain the cross validation procedure.
           Ans.     ross validation is a resampling technique for evaluating machine learning models on a small sample of data.
                   he process includes only one parameter, k, that specifies the number of groups into  hich a given data sample
                  should be divided. It's a popular strategy since it's straightfor ard to grasp and produces a less biased or optimistic
                  estimate of model competence than other approaches, such as a simple train test split.  he follo ing is the general
                  procedure
                  i.   andomly shuffle the dataset.
                  ii.  Organise the data into k groups.
                  iii.   or each distinct group
                     • As a test data set, take a group.
                     •  or training data set, use the remaining groupings.
                     •  it the model to the training set and evaluate it against the test set.

                     •  eep the evaluation score but toss out the model.
                  iv.   sing the sample of model evaluation scores, summarise the model's ability.
              .    hat is  rain  est  plit  valuation   tate the reasons for choosing this technique.
           Ans.     he  train  test  procedure  measures  the  performance  of  machine  learning  algorithms   hen  they  need  to  make
                  predictions on data that  ere not used to train the model.  he technique divides the provided dataset into t o
                  subsets  the training dataset and test dataset.  he reasons for choosing this technique are
                 •  arge dataset
                 •  o estimate the machine learning model's performance on ne  data that  as not used to train the model
                 •  etter computational efficiency
                 • A quick overvie  of model performance

              .     plain the   stages of data preparation.
           Ans.  Stage 4: Data Collection: During the initial data collection phase, data scientists identify available data sources
                   structured, unstructured, and semi structured) relevant to the problem area. If there is a gap in data collection, the
                  data scientist may need to modify data requirements accordingly and collect ne  and or more data.
                 Stage 5: Data Understanding:  After  the  initial  data  collection,  techniques  such  as  descriptive  statistics  and
                 visualisations can be applied to datasets to evaluate the content, quality, and initial insights of the data. Additional
                 data collection may be required to fill the gap.
                      Touchpad Artificial Intelligence (Ver. 2.0)-XII
   139   140   141   142   143   144   145   146   147   148   149