Data Driven decision making final

• Systematic errors


  • Calibration
  • Experimenter drift
  • Sampling bias
• Categories of agricultural data

  • Agronomic
  • Livestock
  • Land
  • Farm management
  • Machine equipment
  • Climate & weather
• Steps of cleaning data

  1. Remove duplicate or wrong data

  2. Fix structural errors

  3. Filter outliers

  4. Handle missing data

  5. Validate data
• Types of error


  • Random
  • Systematic
• Steps for data-driven decisions

  1. Vision

  2. Find data source

  3. Organize data

  4. Perform data analysis

  5. Draw conclusions
• 5 Vs of big data

  • Volume
  • Value
  • Variety
  • Velocity
  • Veracity
• Ethical concerns with data management

  • Privacy
  • Transparency
  • Informed consent
  • Bias
  • Security
  • Data protection
• Core principles of agricultural data transparency

  • Transparency
  • Choice
  • Profitability
  • Security
  • Identity of provider
  • Identity of data partners
  • Disclosure and sale limitations
• Types of statistics


  • Descriptive
  • Inferential
• Data quality domains


  • Utility
  • Objectivity
  • Integrity
• Security solutions


  • Human centric
  • Physical aspect
  • Technology based
• Types of questions to avoid

  • Leading
  • Close ended
  • Vague
• Advantages of precision farming

  • Cost efficient
  • Time
  • Environmental sustainability
  • Data driven decision making
  • Profitability
• Data-driven decision making


  Using facts and data to find patterns, inferences and insights to inform your decision making process
• Quantitative data


  Can be counted or measured
• Qualitative data


  Can be seen, heard, felt, smelled
• Artificial intelligence


  Simulation of human intelligence processes by machines
• ROI


  Return on investment
• Precision farming


  Improving crop yields by using tech or sensors to target exact things
• Agricultural Data Transparent


  Certification based on privacy and security principles for farm data
• Accuracy


  How closely the data reflects the true values (how close your data is to reality)
• Telematics


  Integration of telecommunications to enable all the machinery to be tracked in real time and analyzed together
• SMART


  Specific, measurable, action oriented, relevant, time bound
• Calculating range


  Maximum - minimum
• Calculating mode


  Most frequent number
• Calculating median


  Middle value of data set when arranged ascending to descending
• Ways to measure data reliability


  • Validity
  • Completeness
  • Uniqueness
• Predictor variable


  Independent variable
• Garbage in, garbage out means low quality data means low quality decisions