-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.xml
68 lines (68 loc) · 14.9 KB
/
index.xml
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
<title>Data Science Portfolio on Dr. Eng. Frederick N. Numbisi</title>
<link>https://gohugo-theme-ananke.pages.dev/</link>
<description>Recent content in Data Science Portfolio on Dr. Eng. Frederick N. Numbisi</description>
<generator>Hugo</generator>
<language>en-US</language>
<lastBuildDate>Fri, 12 Apr 2024 11:14:48 -0400</lastBuildDate>
<atom:link href="https://gohugo-theme-ananke.pages.dev/index.xml" rel="self" type="application/rss+xml" />
<item>
<title>project 6: Mapping Arid vegetation communities and habitats</title>
<link>https://gohugo-theme-ananke.pages.dev/post/project-6/</link>
<pubDate>Fri, 12 Apr 2024 11:14:48 -0400</pubDate>
<guid>https://gohugo-theme-ananke.pages.dev/post/project-6/</guid>
<description><ul>
<li>Data for vegetation assemblage and habitat mapping were sourced from three main streams: 1. Inventory data collected during field surveys, 2. Estimated proxies for vegetation structure and phenology – vegetation indices, and 3. Estimates of spectral signatures for ground features from time-series remote sensing imagery.</li>
<li>To map the vegetation assemblages and habitats, the random forest classification algorithm was implemented to delineate clusters of vegetation types, hitherto identified and grouped as clusters of plant assemblages.</li>
<li>Field referenced points and polygons for the vegetation assemblages were used to extract training data from the stack of engineered spatial features (10 explanatory or predictor raster layers).</li>
<li>In selecting a reliable classifier algorithm for the considered vegetation community or habitat types, a Multi-Classifier System (MCS) of algorithms that comprise ensemble modelling, bagging, and boosting: Random Forest ensemble decision tree (RF), Support Vector Machine (SVM), Regularized discriminant analysis (RDA) model, Naïve Bayes classifier (NB), and the Gradient Boosting Model (GBM). All model training was implemented using the “caret” library in R program.</li>
<li>The RF classification algorithm had better performance (overall accuracy and kappa statistics) and was applied to model the spatial location and distribution of the reference classes from field inventory (clusters of vegetation assemblages and communities) and the available spatial (raster image) predictors.</li>
<li>Optimised the RF classifier algorithms using forward feature selection (Meyer et. al., 2018) and spatial cross-validation procedure to reach the best model.</li>
<li>A Leave-Location-Out Cross Validation (LLO-CV) was applied to address potential class selection bias and overfitting, and a resampling control of 10-fold cross-validation with 5 iterations.</li>
<li>The RF model training and cross validation was implemented within the “CAST” library (Meyer et al., 2023) of the R statistical program.</li>
</ul>
<p><figure><img src="https://gohugo-theme-ananke.pages.dev/images/Project_SpatialAnalysisGrid.png"><figcaption>
 <h4>Arid Plant Communities Classifier Grid</h4>
 </figcaption>
</figure>

<figure><img src="https://gohugo-theme-ananke.pages.dev/images/Ashar/WadiAsharVegetation_Figure5.png"><figcaption>
 <h4>Class Uncertainty Distribution</h4>
 </figcaption>
</figure>

<figure><img src="https://gohugo-theme-ananke.pages.dev/images/Ashar/WadiAsharVegetation_Figure3.png"><figcaption>
 <h4>Predicting Area of Applicability (AOA) of Classifier model</h4>
 </figcaption>
</figure>

<figure><img src="https://gohugo-theme-ananke.pages.dev/images/Ashar/WadiAsharVegetation_Figure4.png"><figcaption>
 <h4>Class predictions from different classifier models</h4>
 </figcaption>
</figure>

<figure><img src="https://gohugo-theme-ananke.pages.dev/images/VegetaionCommunitiesMap.png"><figcaption>
 <h4>Vegetation communities and habitat map.</h4>
 </figcaption>
</figure>

<figure><img src="https://gohugo-theme-ananke.pages.dev/images/VegModel_MapBwithAOA.png"><figcaption>
 <h4>Vegetation communities and habitat map with consideration of Area of Applicability (AOA) of classification model.</h4>
 </figcaption>
</figure>
</p></description>
</item>
<item>
<title>project 5: Predictor of Plant distribution clusters</title>
<link>https://gohugo-theme-ananke.pages.dev/post/project-5/</link>
<pubDate>Fri, 14 Apr 2023 11:25:05 -0400</pubDate>
<guid>https://gohugo-theme-ananke.pages.dev/post/project-5/</guid>
<description></description>
</item>
<item>
<title>Project 3: Estimator of Tree Spatial Pattern in Agroforests</title>
<link>https://gohugo-theme-ananke.pages.dev/post/project-3/</link>
<pubDate>Sun, 11 Apr 2021 11:13:32 -0400</pubDate>
<guid>https://gohugo-theme-ananke.pages.dev/post/project-3/</guid>
<description><ul>
<li>Assessed the temporal changes in shade management and spatial dynamics in canopy tree structure and live tree biomass.</li>
<li>Conducted field survey along an age gradient of “family farms” in cocoa agroforests created from forest (fCAFS) and savannah (sCAFS) land cover.</li>
<li>Evaluated the temporal changes in farm structure, relative tree abundance, and live aboveground biomass of the major canopy (shade tree) strata.</li>
<li>Used spatial point process and linear mixed effect analyses to assess contributions of associated perennial trees (AsT) on farm rejuvenation patterns.</li>
<li>Provided insights into farmers’ temporal allocation of uses and prioritization of different tree species associations with age of cocoa agroforests.</li>
<li>Developed recommendations for landscape-specific tree management and considerations in proposing on-farm tree conservation incentives.</li>
</ul>
<p><figure><img src="https://gohugo-theme-ananke.pages.dev/images/SustainabilityFig2.png"><figcaption>
 <h4>Spatial sampling and point pattern analysis steps</h4>
 </figcaption>
</figure>

<figure><img src="https://gohugo-theme-ananke.pages.dev/images/SustainabilityFig9.png"><figcaption>
 <h4>Spatial pattern of cocoa trees across farm type and age gradients</h4>
 </figcaption>
</figure>
</p></description>
</item>
<item>
<title>Project 2: Cocoa Agroforestry Canopy Gap Predictor</title>
<link>https://gohugo-theme-ananke.pages.dev/post/project-2/</link>
<pubDate>Fri, 10 Apr 2020 11:00:59 -0400</pubDate>
<guid>https://gohugo-theme-ananke.pages.dev/post/project-2/</guid>
<description><ul>
<li>Estimated canopy cover distribution using in-situ digital hemispherical photographs (DHPs) sampling and estimates of canopy gap fraction.</li>
<li>Built neural network and random forest regression models of cocoa agroforestry canopy gap fraction and Sentinel-1A SAR backscatter intensity and features.</li>
<li>Developed a combination of different backscatter variables for predicting the canopy gap variability in agroforestry cocoa production landscapes.</li>
<li>Utilise a semi-variogram analysis of canopy gap distribution and spatial clustering distances in different cocoa production landscapes.</li>
<li>Provided new insights into the scale of spatial variability of canopy gaps in relation to farm and landscape management through cocoa agroforestry land use.</li>
<li>Built a proof-of-concept to support development of management tools or strategies on tree inventorying and decisions regarding incentives for shade tree retention and planting in cocoa landscapes.</li>
</ul>
<figure><img src="https://gohugo-theme-ananke.pages.dev/images/DHPprofile_pictures.png"><figcaption>
 <h4>Inventory of canopy cover distribution (ground truth)</h4>
 </figcaption>
</figure>

<figure><img src="https://gohugo-theme-ananke.pages.dev/images/DHP_graphicAbstract.png"><figcaption>
 <h4>Field and spatial data analysis workflow</h4>
 </figcaption>
</figure>

<ul>
<li><a href="https://github.com/Frederick-Numbisi/AgroforestryCanopyGapPrediction">Link to GitHub Repository</a></li>
<li><a href="https://doi.org/10.3390/rs12244163">Link to Reference Scientifica Publication - 2020</a></li>
</ul></description>
</item>
<item>
<title>Project 1: Cocoa Agroforestry Land Use Classifier</title>
<link>https://gohugo-theme-ananke.pages.dev/post/project-1/</link>
<pubDate>Mon, 09 Apr 2018 10:58:08 -0400</pubDate>
<guid>https://gohugo-theme-ananke.pages.dev/post/project-1/</guid>
<description><ul>
<li>Developed a classifier procedure for discriminating perennial cocoa agroforestry land cover.</li>
<li>Compared a multi-spectral optical image from RapidEye, acquired in the dry season, and multi-seasonal C-band SAR of Sentinel 1.</li>
<li>Hand engineered Grey Level Co-occurrence Matrix (GLCM) texture features (images) from Sentinel-1 Image time series covering six dry and four wet seasons from 2015 to 2017.</li>
<li>Optimised Random Forest Ensemble classifier models using different input feature combinations; multi-spectral reflectance, vegetation indices, co-(VV) and cross-(VH) polarised SAR intensity and GLCM texture measures.</li>
<li>Evaluated accuracy metrics and uncertainty Shannon entropy - information loss in classification.</li>
<li>Built a classifier and uncertainty measures that provide reliable validation of class discrimination at different spatial resolution.</li>
</ul>
<figure><img src="https://gohugo-theme-ananke.pages.dev/images/Figure_3_IJGI_NumbisiFN.png"><figcaption>
 <h4>SAR Backscatter Time-series over different land uses</h4>
 </figcaption>
</figure>

<ul>
<li><a href="https://github.com/Frederick-Numbisi/Mapping-Agroforesty-Cocoa-Land-Use">Link to GitHub Repository</a></li>
<li><a href="https://doi.org/10.5194/isprs-archives-XLII-1-339-2018">Link to Reference Scientific Publication 1 - 2018</a></li>
<li><a href="https://doi.org/10.3390/ijgi8040179">Link to Reference Scientific Publication 2 - 2019</a></li>
</ul></description>
</item>
<item>
<title>project 4: Predictor of Important Arid plants hotspots</title>
<link>https://gohugo-theme-ananke.pages.dev/post/project-4/</link>
<pubDate>Thu, 13 Apr 2017 11:15:58 -0400</pubDate>
<guid>https://gohugo-theme-ananke.pages.dev/post/project-4/</guid>
<description><figure><img src="https://gohugo-theme-ananke.pages.dev/images/AlUlaProject_SpatialAnalysisGrid.png"><figcaption>
 <h4>Arid Plant Communities Classifier Grid</h4>
 </figcaption>
</figure></description>
</item>
<item>
<title> About </title>
<link>https://gohugo-theme-ananke.pages.dev/about/</link>
<pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
<guid>https://gohugo-theme-ananke.pages.dev/about/</guid>
<description><figure><img src="https://gohugo-theme-ananke.pages.dev/images/AboutPic_Frederick_cropped.jpg">
</figure>

<p>Remote Sensing and Vegetation Analytics are my passions. I have been working in the nexus of Forestry, Agroforestry, Remote Sensing and Data science fields and conducting vegetation cover inventory and analytics projects for the last 10 years. Transitioning into data science from a natural science background, I have held data management and remote sensing scientist positions in companies ranging from Universities, International NGOs and Key Conservation Research organizations. I strive towards improving my experience and contributing to an understanding of information that intersect between Data Science and Earth Observation. This will leverage different and related goals in vegetation monitoring &amp; management, land use sustainability, and EUDR Due Diligence.</p></description>
</item>
<item>
<title>Contact</title>
<link>https://gohugo-theme-ananke.pages.dev/contact/</link>
<pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
<guid>https://gohugo-theme-ananke.pages.dev/contact/</guid>
<description><p>Follow me on these social media platforms.</p>
<table>
 <thead>
 <tr>
 <th>Platform</th>
 <th>URL</th>
 </tr>
 </thead>
 <tbody>
 <tr>
 <td>Website:</td>
 <td><a href="https://www.fredericknumbisi-dev.org">https://www.fredericknumbisi-dev.org</a></td>
 </tr>
 <tr>
 <td>LinkedIn:</td>
 <td><a href="https://www.linkedin.com/in/frederick-n-numbisi/">https://www.linkedin.com/in/frederick-n-numbisi/</a></td>
 </tr>
 <tr>
 <td>ResearchGate:</td>
 <td><a href="https://www.researchgate.net/profile/Frederick_N_Numbisi">https://www.researchgate.net/profile/Frederick_N_Numbisi</a></td>
 </tr>
 <tr>
 <td>Publons:</td>
 <td><a href="https://publons.com/researcher/3030082/frederick-nkeumoe-numbisi/">https://publons.com/researcher/3030082/frederick-nkeumoe-numbisi/</a></td>
 </tr>
 <tr>
 <td>GitHub:</td>
 <td><a href="https://github.com/Frederick-Numbisi">https://github.com/Frederick-Numbisi</a></td>
 </tr>
 </tbody>
</table></description>
</item>
</channel>
</rss>