Inaugurated by IN-SPACe
ISRO Registered Space Tutor
S7-SA3-0301
What is the Concept of Markov Chains Introduction?
Grade Level:
Class 12
AI/ML, Physics, Biotechnology, FinTech, EVs, Space Technology, Climate Science, Blockchain, Medicine, Engineering, Law, Economics
Definition
What is it?
Markov Chains are mathematical models that describe a sequence of events where the probability of the next event depends only on the current state, and not on any past events. Imagine a chain where each link's position only cares about the link just before it, not the very first link.
Simple Example
Quick Example
Think about whether it will rain tomorrow. If it's raining today, there's a 70% chance it will rain tomorrow and a 30% chance it will be sunny. If it's sunny today, there's a 20% chance it will rain tomorrow and an 80% chance it will be sunny. The weather tomorrow only depends on today's weather, not yesterday's or last week's.
Worked Example
Step-by-Step
Let's say a cricket team can be in two states: 'Winning' (W) or 'Losing' (L). If they are Winning today, there's a 90% chance they Win tomorrow and 10% chance they Lose. If they are Losing today, there's a 40% chance they Win tomorrow and 60% chance they Lose. What is the probability they are Winning on Day 2, if they are Winning on Day 0?
Step 1: Identify the states: W (Winning), L (Losing).
---Step 2: Identify transition probabilities:
- P(W tomorrow | W today) = 0.90
- P(L tomorrow | W today) = 0.10
- P(W tomorrow | L today) = 0.40
- P(L tomorrow | L today) = 0.60
---Step 3: Starting state: Winning on Day 0.
---Step 4: Calculate probability for Day 1:
- P(W on Day 1 | W on Day 0) = 0.90
- P(L on Day 1 | W on Day 0) = 0.10
---Step 5: Calculate probability for Day 2 (they can reach W on Day 2 from W on Day 1 OR L on Day 1):
- P(W on Day 2) = P(W on Day 2 | W on Day 1) * P(W on Day 1) + P(W on Day 2 | L on Day 1) * P(L on Day 1)
- P(W on Day 2) = (0.90 * 0.90) + (0.40 * 0.10)
- P(W on Day 2) = 0.81 + 0.04
---Step 6: Final calculation.
- P(W on Day 2) = 0.85
ANSWER: The probability they are Winning on Day 2 is 0.85 or 85%.
Why It Matters
Markov Chains are powerful tools used in many fields like AI to predict stock prices, in medicine to model disease spread, and in engineering to design reliable systems. Learning this helps you understand how self-driving cars predict traffic or how social media algorithms suggest what you might like next.
Common Mistakes
MISTAKE: Thinking the next state depends on ALL previous states. | CORRECTION: Remember the 'memoryless' property – the next state only depends on the IMMEDIATE current state.
MISTAKE: Confusing probability of being in a state with probability of transitioning to a state. | CORRECTION: 'Probability of being in state X' is different from 'Probability of moving from state Y to state X'. Focus on the arrows (transitions) between states.
MISTAKE: Not ensuring transition probabilities from any state sum up to 1. | CORRECTION: From any given state, the sum of probabilities of transitioning to all possible next states must always equal 1 (or 100%).
Practice Questions
Try It Yourself
QUESTION: A light switch can be 'ON' or 'OFF'. If it's ON, there's a 0.8 chance it stays ON and 0.2 chance it turns OFF. If it's OFF, there's a 0.6 chance it stays OFF and 0.4 chance it turns ON. What is the probability it is ON tomorrow if it is ON today? | ANSWER: 0.8
QUESTION: A student studies for 1 hour, 2 hours, or 3 hours daily. If they study 1 hour today, there's a 0.5 chance they study 1 hour tomorrow, 0.3 chance for 2 hours, and 0.2 chance for 3 hours. If they study 2 hours today, what is the probability they study 1 hour tomorrow? (Assume similar probabilities for other transitions). | ANSWER: We only need the direct transition from 2 hours to 1 hour. If not given, we cannot determine. (Assuming the question implies a direct transition from 2 hours to 1 hour is given or needs to be calculated from full matrix, but for simplicity, the answer is the direct probability if provided. Let's assume P(1 hour tomorrow | 2 hours today) is 0.4. So ANSWER: 0.4)
QUESTION: A local food stall offers 'Dosa' (D) or 'Idli' (I) daily. If they sell Dosa today, there's a 70% chance they sell Dosa tomorrow and 30% chance Idli. If they sell Idli today, there's a 50% chance they sell Dosa tomorrow and 50% chance Idli. If they sold Dosa on Monday, what is the probability they sell Idli on Wednesday? Show steps. | ANSWER: Step 1: Dosa on Monday (Day 0). --- Step 2: Probabilities for Tuesday (Day 1): P(D on Day 1) = 0.7, P(I on Day 1) = 0.3. --- Step 3: Probabilities for Wednesday (Day 2): P(I on Day 2) = P(I on Day 2 | D on Day 1)*P(D on Day 1) + P(I on Day 2 | I on Day 1)*P(I on Day 1). --- P(I on Day 2) = (0.3 * 0.7) + (0.5 * 0.3) = 0.21 + 0.15 = 0.36. --- ANSWER: 0.36 or 36%.
MCQ
Quick Quiz
Which of the following is a key characteristic of a Markov Chain?
The future state depends on all past states.
The probability of the next state depends only on the current state.
It always predicts future events with 100% certainty.
It can only have two possible states.
The Correct Answer Is:
B
Option B correctly describes the 'memoryless' property of Markov Chains, where the next state's probability is solely determined by the current state. Options A, C, and D are incorrect characteristics.
Real World Connection
In the Real World
In India, Markov Chains are used by companies like Ola or Uber to predict taxi demand in different areas at different times, helping drivers know where to go. They also help in predicting customer behavior for online shopping platforms or in analyzing traffic flow in smart cities.
Key Vocabulary
Key Terms
STATE: A specific condition or situation in the system, like 'raining' or 'sunny' | TRANSITION: The movement from one state to another | PROBABILITY: The likelihood of an event happening, expressed as a number between 0 and 1 | MEMORYLESS PROPERTY: The next state only depends on the current state, not the past ones.
What's Next
What to Learn Next
Now that you understand the basics, explore 'Transition Matrices' which are tables used to organize these probabilities. Then, you can dive into 'Steady-State Distributions' to see how systems behave over a very long time.
