AI terms starting with P.

p-value

पी-मान

The probability of seeing results as extreme as observed if the claim were false, by chance.

Padding

प्याडिङ

Adding extra border pixels so a convolution can process an image's edges.

Panoptic Segmentation

प्यानोप्टिक खण्डीकरण

Labeling every pixel by both its object class and its individual instance.

Parameter

प्यारामिटर

A value inside a model (like a weight) learned from data during training.

Part-of-Speech Tagging

शब्द-वर्ग ट्यागिङ

Labeling each word in a sentence as a noun, verb, adjective, and so on.

Pattern Recognition

ढाँचा पहिचान

Identifying regularities or recurring structures in data, a core ability of AI systems.

Percentile

प्रतिशतक

A value below which a given percentage of the data falls.

Perception (AI)

अवबोध / प्रत्यक्षण

How an AI system interprets sensory input to understand its surroundings.

Perceptron

पर्सेप्ट्रोन

The simplest neural unit that combines weighted inputs to produce a single output.

Perplexity

पर्प्लेक्सिटी

A measure of how well a language model predicts text; lower means better prediction.

Persona

पर्सोना

A defined character, tone, and role that an AI assistant adopts in conversation.

Personalization

वैयक्तिकरण

Tailoring content, products, or services to each individual user using AI.

Pilot Project

पाइलट परियोजना

A small initial rollout to test an AI solution before wider deployment.

Pipeline

पाइपलाइन

An automated sequence of steps that moves data through processing, training, and deployment.

Pixel

पिक्सेल

The smallest unit of a digital image, holding color and brightness information.

Planning

योजना निर्माण

Generating a sequence of actions an agent should take to reach a goal.

Policy (RL)

नीति (आरएल)

The strategy an agent uses to decide which action to take in each state.

Pooling

पुलिङ

Reducing the size of feature maps in a CNN while keeping important information.

Pose Estimation

अवस्था अनुमान

Detecting the position and orientation of a person's body or joints in an image.

Positional Encoding

स्थिति सङ्केतन

Adding order information to tokens so a transformer knows their sequence.

Pre-training

पूर्व-तालिम

The initial large-scale training of a model on broad data before any task-specific tuning.

Precision

प्रिसिजन / यथार्थता

Of all the items a model labeled positive, the fraction that were actually positive.

Prediction

भविष्यवाणी / अनुमान

The output a model produces for a given input, such as a label, number, or generated text.

Predictive Analytics

भविष्यसूचक विश्लेषण

Using data and models to forecast future outcomes and trends.

Predictive Maintenance

पूर्वानुमानित मर्मतसम्भार

Using data to predict equipment failures before they happen and service in advance.

Principal Component Analysis (PCA)

प्रमुख घटक विश्लेषण

A technique that reduces many features into a few that capture most of the variation.

Probability

सम्भाव्यता

A number between 0 and 1 expressing how likely an event is to happen.

Procedural Generation

प्रक्रियात्मक सिर्जना

Creating content like levels or textures automatically using algorithms.

Process Mining

प्रक्रिया खनन

Analyzing event logs to discover, monitor, and improve real business processes.

Prompt

प्रम्प्ट

The input text or instruction a user gives to a language model.

Prompt Caching

प्रम्प्ट क्यासिङ

Reusing stored results for repeated prompt parts to cut cost and latency.

Prompt Engineering

प्रम्प्ट इन्जिनियरिङ

The practice of designing effective prompts to get useful outputs from an AI model.

Prompt Injection

प्रम्प्ट इन्जेक्सन

An attack that hides malicious instructions in input to hijack an AI's behavior.

Prompt Template

प्रम्प्ट टेम्प्लेट

A reusable prompt structure with placeholders filled in for each request.

Proof of Concept (PoC)

अवधारणा प्रमाण

A small initial project to test whether an AI idea is feasible before full investment.

Proof of Value (PoV)

मूल्य प्रमाण

A trial showing the concrete business value of an AI solution before scaling.

Provenance

उद्गम / स्रोत-इतिहास

A traceable record of where data or content came from and how it was changed.

Pruning (Neural)

प्रुनिङ (न्युरल)

Removing unimportant weights or neurons to make a model smaller and faster.

PyTorch

पाइटर्च

A popular open-source deep learning framework known for flexibility and ease of use.