Name:- SHUBHAM HANDOO

[Link]- PGP2024025

Q4)

MARKS:- CIRCULAR GLYPS

CHANNELS:- POSITION, SIZE, COLOR, ANGLE

DATASET OF THE VISUAULISATION

YEAR(1995-2008) STATES ( OF USA ) BIM INDEX (OBESE, NORMAL,

OVERWIEGHT)

LIMITATIONS OF THIS DATA:-

1)LACK OF THE EXACT VALUES:- glyps does not contain the exact values in each catepgary

2) Time Presentation: - time line is given but not clear how to choose specific year

3)Color Distinction :- orange & red color will be hard for someone to distinguishg

4)SIZE VARIATION :- sizes of gylps make it hard to compare proportions across sates accurately.

5) OVERLAPPING:- densely populated areas are overlapping

REDESIGN OF THE VISUALISATION: -

Visualization can be choropleth graph with Parula and Vardis color schemes and timeline should be
made bold and clear to identify which year is shown. Color schemes should show values also
corresponding to the
Q3)

1. Marks: Lines are used as the primary mark, connecting data points across multiple parallel
vertical axes.

2. Channels:

 Position: The horizontal position of each vertical axis represents a different attribute
(Region, Month, Product, Sales, Equipment costs, Labor costs).

 Color: While the image is in black and white, line density creates varying shades that
help distinguish between different data points or clusters.

Alternative Improved Visualization: An improved version of this visualization could be an interactive

stacked bar chart with the following key aspects:

1. X-axis: Months of the year

2. Y-axis: Sales volume

3. Stacked bars: Each bar represents total sales for a month, with segments color-coded for
different products

4. Interactive elements:

 Hover tooltips showing detailed information (equipment costs, labor costs) for each
product segment

 Ability to filter by region

 Option to switch between absolute values and percentages

Why this new design is effective:

1. Clarity: It reduces the visual clutter of the parallel coordinates plot, making it easier to spot
trends over time.

2. Hierarchy of information: It prioritizes the most important variables (time, sales, and products)
while making secondary information (costs) available on demand.

3. Comparative analysis: It allows for easy comparison of product performance across months and
overall sales trends.

4. Interactivity: It provides a way to explore the data in depth without overwhelming the user with
all information at once.

5. Familiar format: Most users are accustomed to bar charts, making this visualization more
immediately understandable.

6. Scalability: This design can handle additional data points (e.g., more months or products)
without becoming overly complex.
Q2)

would make a multidimensional filtering for different attributes, and a dashboard for critical health
indicators which are compared to the average score for each indicator, or standard deviation from the
average values on each field. For example: A patient with platelet count 2 standard deviations below the
mean is at a lower risk of clots but at a higher risk of erroneous cuts and infections arising from them.

I would also use all rational available attributes for calculating a health risk score for the patient which I
can plot on a scatterplot against their insurance premium (X Axis: Health Risk, Y axis: Insurance
premium) to see where they lie. We can distribute this scatterplot into quadrants on HIGH/LOW basis
for both insurance premium and health risk, where (LOW, HIGH) would be most ideal for patient
outcomes, and so on.

Alternatively, we can also show a violin graph of different demographics, their insurance premiums and
their risk profiles, we can also show where our patients lies on that.
Q1)

dentifying Suspicious Heavy Payloads

To identify suspicious heavy payload periods in weekly hourly truckload data, follow these steps:

Data Transformation

Hourly Aggregation: Sum truckload weights for each hour across the week.

3-Hour Moving Average: Smooth the data using a sliding window to highlight trends.

Normalization: Apply z-score normalization to detect deviations from the mean, identifying potential
outliers.

Data Visualization

Heatmap: Visualize days of the week on one axis, hours on the other, with color intensity indicating
higher payloads.

Time Series Plot: Chart the 3-hour moving averages, highlighting periods where the z-score exceeds a
threshold (e.g., z > 2) for suspicious activity.

This approach effectively reveals abnormal payload patterns, pinpointing the 3-hour window of
suspicious activity during the week.

Visualizing Potential Black Market Locations

To identify potential smuggling handoff points 20 km from the post:

Location Identification

Road Network Analysis: Use GIS tools to map points 20 km away and focus on intersections, rest stops,
and discreet roads.

Historical Data Integration: Prioritize locations based on known smuggling incidents and traffic patterns.

Visualization

Map Overlay: Use a detailed road network map with markers indicating potential handoff points, color-
coded by risk level.

Heatmap: Layer a heatmap to show areas with higher suspicious activity or traffic density.

Interactive Features: Enable clickable markers showing additional intelligence, like timing of suspected
activities.

This strategy visualizes high-risk locations, optimizing surveillance and resource deployment to intercept
smuggling operations.