Page 266 - AI Ver 3.0 Class 11
P. 266
Problem Scoping
The beginning of any project requires describing the problem, which
is done in the problem-defining phase of the project life cycle. It is a
crucial stage that requires in-depth research of the problem, so that
the desired solution is well-written in an absolute understandable
form. This helps progressively in all the following stages of the project
life cycle, and is also used to trace back any missed out features
that was initially planned to implement in the AI model. Successful
implementation of the AI model is critically dependent on this stage
of the AI life cycle.
Data Collection
Once the problem is identified and defined, we then begin to collect the data for
it. The data that needs to be collected is identified and recorded using machines
that record real-world signals for AI model to work successfully. Data acquisition
is followed by data cleansing, that drops the irrelevant information with respect
to the problem at hand. The inputs are then digitised for processing by the AI
model in place.
Data Exploration
Data exploration serves as a fundamental initial phase in data-driven projects, aimed at deeply understanding the
dataset's structure, characteristics, and underlying insights. This critical process involves summarising essential
statistics such as data types, distributions, and handling missing values,
alongside conducting exploratory data analysis (EDA) to unveil patterns,
relationships, and anomalies. Utilising various visualisation techniques
such as histograms, scatter plots, and box plots enables analysts to
visually grasp data distributions and correlations, thereby guiding
preprocessing decisions like outlier treatment and feature engineering.
By laying this groundwork, data exploration ensures subsequent
analytical and modeling phases are built upon a robust comprehension
of the dataset's intricacies, empowering data-driven decisions with
clarity and reliability.
AI Modelling
This is the phase of implementation of the AI model using suitable
platforms, programming languages, and constructs. An AI model is a
program that has been trained on a set of data to recognise certain
patterns or make certain decisions without further human intervention.
They apply different algorithms to relevant data inputs to achieve the
tasks, or output, they’ve been programmed for. This model is developed
to solve the problem defined in the first phase. The model is then
trained with the training data. The training is iterative in nature, so that the system is prepared for the most unexpected
scenarios. This produces a refined model ready to be evaluated in the upcoming phase.
264 Touchpad Artificial Intelligence (Ver. 3.0)-XI

