Scribd
Upload
58 views1 page

Components of Ai System Design PDF

The components of ai systems degin in pdf format

Uploaded by

Suraj Kumar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
58 views1 page

Components of Ai System Design PDF

The components of ai systems degin in pdf format

Uploaded by

Suraj Kumar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

COMPONENTS OF AI SYSTEM DESIGN

DATA ACQUISITION
Data acquisition is the process of collecting and gathering raw data from various sources to be used for
building machine learning models. This stage is crucial as the quality and relevance of the data directly affect
the success of the model.
Sources: Data can be acquired from databases, APIs, web scraping, sensors, user inputs, publicly available
datasets, or manual entry.
Types: Structured (e.g., SQL databases), unstructured (e.g., text, images), or semi-structured (e.g., JSON,
XML).
Challenges: Incomplete data, high costs, storage requirements, privacy concerns, and real-time data
collection.

DATA PROCESSING AND CLEANING


Once data is acquired, it needs to be processed and cleaned to ensure it is ready for analysis and model
building. This stage involves transforming raw data into a format suitable for model consumption
Cleaning: Handling missing values, removing duplicates, correcting errors, and filtering out noise.
Normalization/Standardization: Scaling data to ensure that variables are on the same scale.
Encoding: Converting categorical variables into numerical values (e.g., one-hot encoding, label encoding).
Outlier Detection: Identifying and possibly removing outliers that can skew model performance.
.

MODEL SELECTION
Model selection involves choosing the most appropriate machine learning algorithm(s) for the task at hand.
Types of Models:
Supervised Learning: Regression (e.g., Linear Regression, Decision Trees) or Classification (e.g., SVM, Random Forest,
Neural Networks).
Unsupervised Learning: Clustering (e.g., K-means, DBSCAN) or Dimensionality Reduction (e.g., PCA).
Reinforcement Learning: For tasks where the model learns via trial and error, such as robotics or game playing.

TRAINING AND EVALUATION


Once the model is selected, it needs to be trained on the dataset and evaluated to ensure it performs well on unseen data.
Training: The model learns from the data by finding patterns and adjusting its internal parameters.
Training Set: A portion of the data is used to train the model.
Optimization Algorithms: Methods like Gradient Descent help in tuning model parameters.
Evaluation: Assessing the performance of the model.
Validation Set: A separate dataset used to tune the hyperparameters and prevent overfitting.
Metrics: Depending on the problem, different evaluation metrics are used:
Classification: Accuracy, Precision, Recall, F1-Score, AUC-ROC.
Regression: Mean Absolute Error (MAE), Root Mean Squared Error (RMSE).
Cross-validation: Splitting the dataset into k-folds to ensure the model generalizes well.

DEPLOYMENT AND MONITORING


Once a model has been trained and evaluated, the final step is to deploy it in a real-world environment
and monitor its performance over time.
Deployment: Involves making the model available for use in production systems.
APIs: The model can be wrapped in an API to make predictions in real-time.
Batch Processing: Running predictions periodically on new data.
Scalability: Ensuring the model can handle production workloads (e.g., deploying on cloud
platforms).
Monitoring: Tracking the model's performance to ensure it continues to deliver accurate predictions.
Model Drift: Over time, the data may change, leading to decreased model accuracy.
Performance Tracking: Regularly evaluating the model's performance using the same metrics.

Name: Uttam Prabhakar Kamble


Class: T.Y.BCA(Science) Div: B
Roll no: 84

You might also like